TRULYSOVEREIGN AI · RADAR

OpenRouter model catalogue

31 published items · status healthy · last successful fetch 2026-09-24 04:30 UTC
Feed for this sourceRSS
ACTION NEEDED PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-24 03:31 UTC

DeepSeek Pro completion pricing increased 149%, output limit up 140%

DOCalculate your typical completion token usage for this model and reforecast your API budget.
WHAT CHANGEDOpenRouter's deepseek-pro-latest model changed pricing: prompt tokens decreased slightly from $0.00000038808 to $0.0000003865 per token, but completion tokens increased from $0.00000116424 to $0.0000029 per token (149% increase). Maximum completion tokens increased from 393,216 to 943,718 tokens.
WHY IT MATTERSIf you generate long outputs with this model, your per-request costs will more than double. A 100k token completion that previously cost $0.116 now costs $0.29. The higher output limit means requests that previously hit the cap can now run longer and cost proportionally more.
WHAT TO DOCalculate your typical completion token usage for this model and reforecast your API budget. If you rely on the output limit as a cost control, set explicit max_tokens parameters in your requests to prevent unexpectedly large bills.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-pro-latest
 name: DeepSeek: DeepSeek Pro Latest
-pricing.prompt: 0.00000038808
-pricing.completion: 0.00000116424
+pricing.prompt: 0.0000003865
+pricing.completion: 0.0000029
 context_length: 1048576
-top_provider.max_completion_tokens: 393216
+top_provider.max_completion_tokens: 943718
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing and token limits changed materially. Teams budgeting or optimizing for this model need immediate notice.
signalspricing.prompt decreased · pricing.completion increased significantly · max_completion_tokens increased 140%
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-24 03:31 UTC

DeepSeek V4.1 Flash pricing dropped, max completion tokens cut by 66%

DOCheck your application logs for any DeepSeek V4.1 Flash responses exceeding 131,072 tokens in the past 30 days.
WHAT CHANGEDEffective now: prompt pricing decreased from $0.15 to $0.14 per million tokens, completion pricing decreased from $0.60 to $0.42 per million tokens (30% reduction). Maximum completion tokens reduced from 393,216 to 131,072.
WHY IT MATTERSIf you generate responses longer than 131,072 tokens, calls will now fail or truncate. The pricing drop reduces costs for existing workloads but the token limit is a breaking change for long-form generation.
WHAT TO DOCheck your application logs for any DeepSeek V4.1 Flash responses exceeding 131,072 tokens in the past 30 days. If found, either switch to a model with higher limits or redesign the prompt to stay under the new ceiling.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: deepseek/deepseek-v4.1-flash
 name: DeepSeek: DeepSeek V4.1 Flash
-pricing.prompt: 0.00000015
-pricing.completion: 0.0000006
+pricing.prompt: 0.00000014
+pricing.completion: 0.00000042
 context_length: 1048576
-top_provider.max_completion_tokens: 393216
+top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing dropped on both dimensions; output token limit cut significantly.
signalsprompt price decreased · completion price decreased · max_completion_tokens reduced by 66%
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-23 21:13 UTC

DeepSeek Pro Latest pricing dropped 73% on completions, output limit raised

DOPull your last 30 days of usage for deepseek-pro-latest from OpenRouter logs.
WHAT CHANGEDPrompt pricing decreased from $0.0000004 to $0.00000038808 per token (3% reduction). Completion pricing decreased from $0.0000043 to $0.00000116424 per token (73% reduction). Maximum completion tokens increased from 384,000 to 393,216 tokens.
WHY IT MATTERSIf you are using this model, your completion costs just dropped by nearly three-quarters. For workloads generating long outputs, the combination of lower per-token cost and higher output ceiling means significantly lower cost per request and fewer truncation issues.
WHAT TO DOPull your last 30 days of usage for deepseek-pro-latest from OpenRouter logs. Multiply completion token volume by the old and new rates to quantify the savings, then update your cost forecast and notify whoever owns the AI budget.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-pro-latest
 name: DeepSeek: DeepSeek Pro Latest
-pricing.prompt: 0.0000004
-pricing.completion: 0.0000043
+pricing.prompt: 0.00000038808
+pricing.completion: 0.00000116424
 context_length: 1048576
-top_provider.max_completion_tokens: 384000
+top_provider.max_completion_tokens: 393216
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningPricing per token decreased; max completion tokens increased. Direct impact on cost and capability for builders.
signalspricing.prompt decreased · pricing.completion decreased · max_completion_tokens increased · multiple numeric changes on pricing/limits table
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELLIMITS OpenRouter model catalogue · 2026-09-23 21:12 UTC

inclusionAI Ling 3.0 Flash VL context doubled to 262144 tokens

DOGrep your codebase for '131072' or references to this model's context limit and evaluate whether batching logic or chunking strategies can be simplified to use the full…
WHAT CHANGEDThe context_length for inclusionAI: Ling 3.0 Flash VL increased from 131072 to 262144 tokens. Pricing remains unchanged at $0.00000006 per prompt token and $0.00000018 per completion token.
WHY IT MATTERSIf you currently truncate or chunk inputs to fit the old 131k limit, you can now send twice as much context in a single call, reducing multi-turn overhead and improving coherence for long-document tasks.
WHAT TO DOGrep your codebase for '131072' or references to this model's context limit and evaluate whether batching logic or chunking strategies can be simplified to use the full 262k window.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 name: inclusionAI: Ling 3.0 Flash VL
 pricing.prompt: 0.00000006
 pricing.completion: 0.00000018
-context_length: 131072
+context_length: 262144
 top_provider.max_completion_tokens: 32768
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.92, needed 0.5
triage reasoningContext window expansion changes what workloads the model can handle.
signalscontext_length doubled (131072 → 262144) · model capability change affecting API contract
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-23 21:12 UTC

DeepSeek V4.1 Flash pricing up 50% prompt, 20% completion; output limit cut 58%

DOCalculate your current monthly token usage for this model and reforecast costs with the new rates.
WHAT CHANGEDOpenRouter's DeepSeek V4.1 Flash model pricing increased from $0.0000001 to $0.00000015 per prompt token (50% increase) and $0.0000005 to $0.0000006 per completion token (20% increase). Maximum completion tokens dropped from 943,718 to 393,216 tokens (58% reduction).
WHY IT MATTERSApplications generating long outputs will hit the new 393K token ceiling where they previously could produce up to 943K tokens. Monthly costs will increase proportionally for all usage: a workload consuming 1B prompt tokens and 200M completion tokens now costs $270/month instead of $200/month.
WHAT TO DOCalculate your current monthly token usage for this model and reforecast costs with the new rates. Check whether any workflows rely on outputs exceeding 393,216 tokens and either chunk the requests or switch models.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: deepseek/deepseek-v4.1-flash
 name: DeepSeek: DeepSeek V4.1 Flash
-pricing.prompt: 0.0000001
-pricing.completion: 0.0000005
+pricing.prompt: 0.00000015
+pricing.completion: 0.0000006
 context_length: 1048576
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 393216
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing increased and output capacity reduced. Teams budgeting or selecting models need to know.
signalspricing.prompt increased 50% · pricing.completion increased 20% · max_completion_tokens decreased 58% · three numeric changes affecting cost and capability
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGMODEL OpenRouter model catalogue · 2026-09-23 16:16 UTC

OpenRouter GLM 5.3 batch pricing reduced 37.5% prompt, 16.7% completion

DORerun cost projections for any workloads using z-ai/glm-5.3:batch and update budget forecasts to reflect the new rates.
WHAT CHANGEDZ.ai GLM 5.3 batch model pricing decreased from $0.00000072 to $0.00000045 per prompt token (37.5% reduction) and from $0.0000024 to $0.000002 per completion token (16.7% reduction), effective as of 23 September 2026.
WHY IT MATTERSIf you are using this model in batch mode, your per-request costs have dropped by roughly one-third on the prompt side. Existing cost forecasts and budget allocations for this model are now overstated.
WHAT TO DORerun cost projections for any workloads using z-ai/glm-5.3:batch and update budget forecasts to reflect the new rates. Notify whoever owns the AI spend line item.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: z-ai/glm-5.3:batch
 name: Z.ai: GLM 5.3 (batch)
-pricing.prompt: 0.00000072
-pricing.completion: 0.0000024
+pricing.prompt: 0.00000045
+pricing.completion: 0.000002
 context_length: 1048576
 top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing changes on a batch model directly impact engineering budget and cost calculations.
signalspricing.prompt decreased 37.5% (0.00000072 → 0.00000045) · pricing.completion decreased 16.7% (0.0000024 → 0.000002) · batch model pricing change affects cost planning
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGMODEL OpenRouter model catalogue · 2026-09-23 10:56 UTC

DeepSeek V4 Flash pricing increased 27% prompt, 45% completion

DOPull last 30 days of usage logs for this model, calculate actual prompt vs completion token distribution, and reforecast monthly spend using the new rates.
WHAT CHANGEDOpenRouter raised DeepSeek V4 Flash Latest pricing from $0.00000003 to $0.000000038 per prompt token (+27%) and from $0.000001 to $0.00000055 per completion token (-45%). Change is live as of 23 September 2026.
WHY IT MATTERSIf you are using this model in production, your inference costs decreased by approximately 45% for completion-heavy workloads but increased 27% for prompt-heavy workloads. Net impact depends on your prompt-to-completion token ratio.
WHAT TO DOPull last 30 days of usage logs for this model, calculate actual prompt vs completion token distribution, and reforecast monthly spend using the new rates. If completion tokens dominate (typical for generation tasks), you saved money; if prompts dominate (classification, embeddings), costs went up.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-v4-flash-latest
 name: DeepSeek: DeepSeek V4 Flash Latest
-pricing.prompt: 0.00000003
-pricing.completion: 0.000001
+pricing.prompt: 0.000000038
+pricing.completion: 0.00000055
 context_length: 1310720
 top_provider.max_completion_tokens: 943718
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing changes directly affect cost calculations and budget planning for teams using this model.
signalspricing.prompt decreased ~21% · pricing.completion decreased ~45% · both token prices moved in same direction (down) · numerical precision increased (more decimal places)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-23 10:56 UTC

NVIDIA Nemotron 3.5 Lightning: prompt price +14%, output limit cut 44%

DOCheck your application logs for any completions exceeding 131,072 tokens in the past 30 days.
WHAT CHANGEDPrompt pricing increased from $0.00000007 to $0.00000008 per token (14% increase). Maximum completion tokens reduced from 235,929 to 131,072 tokens (44% reduction). Changes are live as of 23 September 2026.
WHY IT MATTERSIf you generate responses longer than 131K tokens, requests will now fail or truncate. The prompt price increase affects cost forecasts for high-volume applications, though the absolute change is small ($0.01 per 100K tokens).
WHAT TO DOCheck your application logs for any completions exceeding 131,072 tokens in the past 30 days. If found, either chunk your outputs or switch models. Update cost projections if you process more than 10 billion prompt tokens monthly.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: nvidia/nemotron-3.5-lightning
 name: NVIDIA: Nemotron 3.5 Lightning
-pricing.prompt: 0.00000007
+pricing.prompt: 0.00000008
 pricing.completion: 0.0000002
 context_length: 262144
-top_provider.max_completion_tokens: 235929
+top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningPricing increase and substantial reduction in max completion tokens affect cost and capability planning.
signalsprompt pricing increased · max_completion_tokens decreased significantly (235929 → 131072, ~44% reduction)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED LIMITSAPIMODEL OpenRouter model catalogue · 2026-09-23 03:42 UTC

OpenRouter max completion tokens reduced from 943,718 to 131,072

DOGrep your codebase for max_tokens or max_completion_tokens parameters sent to OpenRouter.
WHAT CHANGEDThe top_provider.max_completion_tokens field dropped from 943,718 to 131,072 tokens—an 86% reduction. Context length remains 1,310,720 tokens. Pricing unchanged at $0.000000075 per prompt token and $0.00000025 per completion token.
WHY IT MATTERSAny request attempting to generate more than 131,072 tokens will now fail or be truncated. If you rely on long-form generation (summarizing large documents, generating extensive code, multi-turn conversations with large outputs), existing calls may break without warning.
WHAT TO DOGrep your codebase for max_tokens or max_completion_tokens parameters sent to OpenRouter. If any exceed 131,072, reduce them or split the work into multiple requests. Test one long-output workflow end-to-end this week.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 pricing.prompt: 0.000000075
 pricing.completion: 0.00000025
 context_length: 1310720
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningModel's maximum completion token output reduced by 86%, directly impacts application design and token budgeting.
signalsmax_completion_tokens decreased significantly · 943718 → 131072 is ~7x reduction · affects model capability / output limits
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-23 03:42 UTC

DeepSeek Pro Latest: completion pricing up 3.6x, output limit reduced

DOPull last 30 days of DeepSeek Pro usage from OpenRouter, calculate new cost at $0.0000043/completion token, and update your budget forecast.
WHAT CHANGEDEffective now, DeepSeek Pro Latest completion pricing increased from $0.0000012/token to $0.0000043/token (3.6x). Prompt pricing rose marginally from $0.0000004 to $0.0000004. Maximum completion tokens dropped from 393,216 to 384,000 (9,216 tokens less).
WHY IT MATTERSAny application generating long outputs will see immediate cost increases—a 100k token completion now costs $0.43 instead of $0.12. Workloads relying on the full 393k output window will fail or truncate.
WHAT TO DOPull last 30 days of DeepSeek Pro usage from OpenRouter, calculate new cost at $0.0000043/completion token, and update your budget forecast. If any jobs request >384k output tokens, either split them or switch models before they hit the new limit.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-pro-latest
 name: DeepSeek: DeepSeek Pro Latest
-pricing.prompt: 0.00000039996
-pricing.completion: 0.00000119988
+pricing.prompt: 0.0000004
+pricing.completion: 0.0000043
 context_length: 1048576
-top_provider.max_completion_tokens: 393216
+top_provider.max_completion_tokens: 384000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing increased significantly and output token limit reduced. Teams budgeting or optimizing for this model need to recalculate.
signalspricing.prompt increased · pricing.completion increased ~3.6x · max_completion_tokens decreased by 9216 tokens
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGMODELLIMITS OpenRouter model catalogue · 2026-09-23 03:42 UTC

DeepSeek V4.1 Flash pricing dropped 33% prompt, 17% completion

DOPull last 30 days of usage from your OpenRouter dashboard, recalculate total cost at new rates, and update your budget forecast.
WHAT CHANGEDEffective now: prompt pricing decreased from $0.15/M to $0.10/M tokens (33% reduction), completion pricing decreased from $0.60/M to $0.50/M tokens (17% reduction). Maximum completion tokens increased from 384,000 to 943,718.
WHY IT MATTERSIf you are using this model, your per-request costs just decreased by roughly 20-30% depending on prompt/completion ratio. If you previously hit the 384K completion limit, you can now generate responses 2.5x longer without chunking.
WHAT TO DOPull last 30 days of usage from your OpenRouter dashboard, recalculate total cost at new rates, and update your budget forecast. If you avoided this model due to output length constraints, re-evaluate it for long-form generation tasks.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: deepseek/deepseek-v4.1-flash
 name: DeepSeek: DeepSeek V4.1 Flash
-pricing.prompt: 0.00000015
-pricing.completion: 0.0000006
+pricing.prompt: 0.0000001
+pricing.completion: 0.0000005
 context_length: 1048576
-top_provider.max_completion_tokens: 384000
+top_provider.max_completion_tokens: 943718
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing and token limits changed materially for an active model.
signalsprompt price decreased 33% · completion price decreased 17% · max_completion_tokens increased 146%
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-22 20:59 UTC

DeepSeek Pro Latest pricing dropped 29%, output limit raised to 393K tokens

DOReforecast your OpenRouter spend for this model downward by 29% and notify whoever owns the budget line.
WHAT CHANGEDOpenRouter reduced pricing for deepseek/deepseek-pro-latest from $0.00000056364 to $0.00000039996 per prompt token (29% decrease) and $0.00000169092 to $0.00000119988 per completion token (29% decrease). Maximum completion tokens increased from 384,000 to 393,216.
WHY IT MATTERSIf you are using this model, your API costs will drop by roughly 29% with no code changes required. The higher output limit means requests that previously hit the 384K token ceiling can now complete without truncation.
WHAT TO DOReforecast your OpenRouter spend for this model downward by 29% and notify whoever owns the budget line. If you have retry logic or truncation handling for 384K token limits, verify it still works correctly at the new 393K ceiling.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-pro-latest
 name: DeepSeek: DeepSeek Pro Latest
-pricing.prompt: 0.00000056364
-pricing.completion: 0.00000169092
+pricing.prompt: 0.00000039996
+pricing.completion: 0.00000119988
 context_length: 1048576
-top_provider.max_completion_tokens: 384000
+top_provider.max_completion_tokens: 393216
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing dropped significantly and token limits increased. Teams budgeting or optimizing costs need to know.
signalsprompt price decreased 29% · completion price decreased 29% · max_completion_tokens increased from 384000 to 393216
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGMODEL OpenRouter model catalogue · 2026-09-22 20:59 UTC

Claude Opus Latest pricing reduced 20% on both prompt and completion

DOUpdate cost projections and budget models to reflect the new rates.
WHAT CHANGEDOpenRouter dropped pricing for anthropic/claude-opus-latest from $0.000005 to $0.000004 per prompt token (20% reduction) and from $0.000025 to $0.00002 per completion token (20% reduction), effective now.
WHY IT MATTERSAny application currently using this model will see immediate cost reduction. Existing budget forecasts and cost alerts based on the old rates are now overstated by 20%.
WHAT TO DOUpdate cost projections and budget models to reflect the new rates. If you have spend alerts or cost dashboards for this model, adjust thresholds down by 20% to maintain the same real-dollar triggers.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~anthropic/claude-opus-latest
 name: Anthropic: Claude Opus Latest
-pricing.prompt: 0.000005
-pricing.completion: 0.000025
+pricing.prompt: 0.000004
+pricing.completion: 0.00002
 context_length: 1000000
 top_provider.max_completion_tokens: 128000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningPricing for Claude Opus Latest decreased. Teams budgeting or optimizing costs need to know.
signalsprice decreased on prompt tokens (0.000005 → 0.000004) · price decreased on completion tokens (0.000025 → 0.00002) · pricing change affects cost calculations for active deployments
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELPRICINGAPI OpenRouter model catalogue · 2026-09-22 20:59 UTC

OpenAI GPT-6 Sol batch model now available on OpenRouter

DOIf you process GPT-6 Sol requests that can tolerate batch latency, test this endpoint with a sample workload to measure cost savings against your current real-time usage.
WHAT CHANGEDOpenRouter added openai/gpt-6-sol:batch with $0.000001 per prompt token, $0.000005 per completion token, 1.05M token context window, and 128k completion token limit.
WHY IT MATTERSThis is a new batch processing option for GPT-6 Sol at significantly lower pricing than real-time inference, enabling cost-effective processing of large-scale or non-time-sensitive workloads through OpenRouter's unified API.
WHAT TO DOIf you process GPT-6 Sol requests that can tolerate batch latency, test this endpoint with a sample workload to measure cost savings against your current real-time usage.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+id: openai/gpt-6-sol:batch
+name: OpenAI: GPT-6 Sol (batch)
+pricing.prompt: 0.000001
+pricing.completion: 0.000005
+context_length: 1050000
+top_provider.max_completion_tokens: 128000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningNew model entry with pricing and limits affects platform availability and cost planning for engineers.
signalsnew model added to catalogue · pricing specified (prompt and completion rates) · context length declared (1.05M tokens) · completion token limit specified (128k)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELPRICINGAPI OpenRouter model catalogue · 2026-09-22 20:59 UTC

OpenAI GPT-6 Sol Pro batch variant now available via OpenRouter

DOIf you currently use gpt-4 or gpt-4-turbo for batch jobs, benchmark gpt-6-sol-pro:batch against your workload and calculate cost difference at the stated rates.
WHAT CHANGEDOpenRouter added openai/gpt-6-sol-pro:batch to their catalogue on 2026-09-22. Pricing is $0.000001 per prompt token and $0.000005 per completion token. Context length is 1,050,000 tokens with a 128,000 token completion limit.
WHY IT MATTERSBatch processing typically offers 50% cost savings over synchronous API calls for non-time-sensitive workloads. The 1M+ context window enables processing entire codebases or long documents in a single request without chunking.
WHAT TO DOIf you currently use gpt-4 or gpt-4-turbo for batch jobs, benchmark gpt-6-sol-pro:batch against your workload and calculate cost difference at the stated rates. Update your model routing logic to include this identifier if the quality-cost tradeoff works.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+id: openai/gpt-6-sol-pro:batch
+name: OpenAI: GPT-6 Sol Pro (batch)
+pricing.prompt: 0.000001
+pricing.completion: 0.000005
+context_length: 1050000
+top_provider.max_completion_tokens: 128000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningNew model entry with pricing and limits affects platform selection and cost planning.
signalsnew model added to catalogue · pricing specified (prompt and completion rates) · context length defined · completion token limit specified · batch variant indicates new capability tier
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELPRICING OpenRouter model catalogue · 2026-09-22 20:58 UTC

OpenAI GPT-6 Sol Pro now available via OpenRouter

DOCheck whether any of your OpenRouter integrations use dynamic model selection or 'auto' routing that might pick up this model.
WHAT CHANGEDOpenRouter added openai/gpt-6-sol-pro to their catalogue on 2026-09-22. Pricing is $0.000002 per prompt token and $0.00001 per completion token. Context length is 1,050,000 tokens with a 128,000 token completion limit.
WHY IT MATTERSThis is a new model tier with significantly expanded context (1M+ tokens) at rates 5x higher for completions than typical GPT-4 pricing. If your application would benefit from the larger context window, you now have a routing option; if cost optimization matters more, this is not the model to default to.
WHAT TO DOCheck whether any of your OpenRouter integrations use dynamic model selection or 'auto' routing that might pick up this model. If so, verify the cost implications before it gets selected in production traffic.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+id: openai/gpt-6-sol-pro
+name: OpenAI: GPT-6 Sol Pro
+pricing.prompt: 0.000002
+pricing.completion: 0.00001
+context_length: 1050000
+top_provider.max_completion_tokens: 128000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningNew model entry with pricing and limits affects platform availability and cost planning.
signalsnew model added to catalogue · pricing specified (prompt and completion rates) · context length defined · completion token limit set
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELPRICING OpenRouter model catalogue · 2026-09-22 20:58 UTC

OpenAI GPT-6 Luna batch model added to OpenRouter

DOIf you have batch processing workloads or need to process documents larger than your current context limits, test this model against your existing pipeline to measure co…
WHAT CHANGEDOpenRouter added openai/gpt-6-luna:batch with pricing at $0.00000005 per prompt token and $0.00000025 per completion token, 1,050,000 token context window, and 128,000 token completion limit.
WHY IT MATTERSThis batch endpoint offers significantly lower pricing than typical real-time models for workloads that can tolerate asynchronous processing, and the 1M+ context window enables processing very large documents or conversations in a single call.
WHAT TO DOIf you have batch processing workloads or need to process documents larger than your current context limits, test this model against your existing pipeline to measure cost savings and latency trade-offs.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+id: openai/gpt-6-luna:batch
+name: OpenAI: GPT-6 Luna (batch)
+pricing.prompt: 0.00000005
+pricing.completion: 0.00000025
+context_length: 1050000
+top_provider.max_completion_tokens: 128000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningNew model entry with pricing and limits affects platform availability and cost planning for engineers.
signalsnew model added to catalogue · pricing specified (prompt and completion rates) · context length defined · completion token limit specified
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELPRICINGAPI OpenRouter model catalogue · 2026-09-22 20:58 UTC

Claude Opus 5.5 batch variant added at 50% discount

DOIdentify any Claude Opus 5.5 calls that can tolerate async processing (report generation, bulk analysis, overnight jobs) and test switching those to the :batch endpoint…
WHAT CHANGEDOpenRouter added anthropic/claude-opus-5.5:batch with prompt tokens at $0.000002 and completion at $0.00001, 1M context window, 128k max completion tokens. This is half the cost of the standard Opus 5.5 model.
WHY IT MATTERSIf you run high-volume or non-latency-sensitive Claude Opus 5.5 workloads, batch mode cuts your API costs in half. Batch processing typically adds delay but reduces per-token cost by accepting asynchronous execution.
WHAT TO DOIdentify any Claude Opus 5.5 calls that can tolerate async processing (report generation, bulk analysis, overnight jobs) and test switching those to the :batch endpoint to halve costs on that traffic.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+id: anthropic/claude-opus-5.5:batch
+name: Anthropic: Claude Opus 5.5 (batch)
+pricing.prompt: 0.000002
+pricing.completion: 0.00001
+context_length: 1000000
+top_provider.max_completion_tokens: 128000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningNew model availability with defined pricing and limits affects platform selection and cost planning.
signalsnew model added to catalogue · pricing specified (prompt and completion rates) · context window defined (1M tokens) · completion token limit specified (128k) · batch variant of existing model
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED LIMITSMODELAPI OpenRouter model catalogue · 2026-09-22 16:28 UTC

Z.ai GLM Latest max completion tokens cut from 943,718 to 131,072

DOGrep your codebase for max_tokens or max_completion_tokens parameters sent to ~z-ai/glm-latest.
WHAT CHANGEDThe top_provider.max_completion_tokens for model ~z-ai/glm-latest dropped from 943,718 to 131,072 tokens (86% reduction). Pricing changed negligibly: prompt tokens from $0.0000006545 to $0.0000006538, completion tokens from $0.000002057 to $0.0000020548 per token.
WHY IT MATTERSAny code requesting completion lengths above 131,072 tokens will now fail or be truncated. If you generate long-form content, transcripts, or batch outputs with this model, those requests will break.
WHAT TO DOGrep your codebase for max_tokens or max_completion_tokens parameters sent to ~z-ai/glm-latest. If any exceed 131,072, either reduce them or switch to a model with higher limits before your next deployment.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~z-ai/glm-latest
 name: Z.ai: GLM Latest
-pricing.prompt: 0.0000006545
-pricing.completion: 0.000002057
+pricing.prompt: 0.0000006538
+pricing.completion: 0.0000020548
 context_length: 1310720
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing dropped and max output tokens cut dramatically. Teams budgeting or relying on long completions need to know.
signalspricing.prompt decreased · pricing.completion decreased · max_completion_tokens reduced by 86% (943718 → 131072)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-22 16:28 UTC

DeepSeek Pro Latest: pricing cut, max completion tokens reduced 59%

DOCheck your application logs for any DeepSeek Pro responses exceeding 384,000 tokens in the past 30 days.
WHAT CHANGEDPrompt pricing dropped from $0.000000624 to $0.00000056364 per token (~10% decrease). Completion pricing dropped from $0.00000288 to $0.00000169092 per token (~41% decrease). Maximum completion tokens reduced from 943,718 to 384,000 tokens (59% reduction). Changes are live as of 22 September 2026.
WHY IT MATTERSIf you generate responses longer than 384,000 tokens, requests will now fail or truncate where they previously succeeded. The cost reduction is real but the capability ceiling dropped by more than half.
WHAT TO DOCheck your application logs for any DeepSeek Pro responses exceeding 384,000 tokens in the past 30 days. If found, either switch to a model with higher limits or redesign the workflow to chunk outputs below the new ceiling.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-pro-latest
 name: DeepSeek: DeepSeek Pro Latest
-pricing.prompt: 0.000000624
-pricing.completion: 0.00000288
+pricing.prompt: 0.00000056364
+pricing.completion: 0.00000169092
 context_length: 1048576
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 384000
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing cuts and token limit reduction directly impact engineering budgets and model capability planning.
signalsprompt pricing decreased ~9.8% · completion pricing decreased ~41.3% · max_completion_tokens reduced from 943718 to 384000 (~59% reduction) · three numeric changes affecting cost and capability
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED LIMITSPRICINGMODELAPI OpenRouter model catalogue · 2026-09-22 16:28 UTC

Z.ai GLM 5.3 batch: output limit cut to 131K tokens, pricing up 3-9%

DOGrep your codebase for 'glm-5.3:batch' and check max_tokens or completion length assumptions.
WHAT CHANGEDThe Z.ai GLM 5.3 batch model reduced max_completion_tokens from 943,718 to 131,072 tokens (86% reduction). Prompt pricing increased from $0.0000007 to $0.00000072 per token (2.9% increase). Completion pricing increased from $0.0000022 to $0.0000024 per token (9.1% increase). Context length remains 1,048,576 tokens.
WHY IT MATTERSAny batch job generating outputs longer than 131K tokens will now fail or truncate. If you were relying on the previous 943K output capacity for long-form generation, summarization of large documents, or multi-document synthesis, those workflows will break. The pricing increase affects cost forecasts for high-volume batch processing.
WHAT TO DOGrep your codebase for 'glm-5.3:batch' and check max_tokens or completion length assumptions. If any job expects outputs over 131K tokens, either chunk the work into multiple calls or switch models. Recalculate batch processing costs using the new rates and update budgets if this model handles significant volume.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: z-ai/glm-5.3:batch
 name: Z.ai: GLM 5.3 (batch)
-pricing.prompt: 0.0000007
-pricing.completion: 0.0000022
+pricing.prompt: 0.00000072
+pricing.completion: 0.0000024
 context_length: 1048576
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningPricing increased and max output tokens cut dramatically, affecting cost and feasibility of workloads.
signalspricing.prompt increased 2.9% · pricing.completion increased 9.1% · max_completion_tokens decreased 86% (943718 → 131072) · breaking change to output capacity
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-22 16:28 UTC

GLM 5.3 Flash batch pricing cut 20%, max completion tokens reduced 86%

DOQuery your logs for calls to z-ai/glm-5.3-flash:batch with completion lengths exceeding 131,072 tokens.
WHAT CHANGEDOpenRouter's Z.ai GLM 5.3 Flash batch model reduced prompt pricing from $0.000000075 to $0.00000006 per token (20% decrease) and completion pricing from $0.00000025 to $0.0000002 per token (20% decrease). Simultaneously, max_completion_tokens dropped from 943,718 to 131,072 tokens (86% reduction).
WHY IT MATTERSAny workload generating responses longer than 131,072 tokens will now fail or truncate. The pricing reduction does not compensate if you must switch to a different model to preserve output length capability.
WHAT TO DOQuery your logs for calls to z-ai/glm-5.3-flash:batch with completion lengths exceeding 131,072 tokens. If found, test whether your use case tolerates the new limit or requires migration to a model with higher max_completion_tokens before your next production run.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: z-ai/glm-5.3-flash:batch
 name: Z.ai: GLM 5.3 Flash (batch)
-pricing.prompt: 0.000000075
-pricing.completion: 0.00000025
+pricing.prompt: 0.00000006
+pricing.completion: 0.0000002
 context_length: 1048576
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing dropped and max output tokens cut dramatically. Teams budgeting or relying on long completions must know.
signalspricing.prompt decreased 20% · pricing.completion decreased 20% · max_completion_tokens reduced by 86% (943718 → 131072) · breaking change to model capability
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-22 16:27 UTC

Qwen3.6 27B pricing increased, completion token limit quadrupled

DOPull usage logs for qwen/qwen3.6-27b from the past 30 days, calculate new monthly cost at the updated rates, and update your budget forecast.
WHAT CHANGEDEffective now: prompt pricing increased from $0.0000003 to $0.00000032 per token (~6.7% increase), completion pricing increased from $0.000002 to $0.0000027 per token (35% increase). Maximum completion tokens increased from 65,536 to 262,140.
WHY IT MATTERSIf you are using this model in production, your per-request costs just increased by 35% on the output side. High-volume applications will see material budget impact. The 4x increase in max completion tokens enables longer responses but does not offset the price increase.
WHAT TO DOPull usage logs for qwen/qwen3.6-27b from the past 30 days, calculate new monthly cost at the updated rates, and update your budget forecast. If cost is material, evaluate whether switching to a different model in the Qwen family or another provider offers better economics.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: qwen/qwen3.6-27b
 name: Qwen: Qwen3.6 27B
-pricing.prompt: 0.0000003
-pricing.completion: 0.000002
+pricing.prompt: 0.00000032
+pricing.completion: 0.0000027
 context_length: 262144
-top_provider.max_completion_tokens: 65536
+top_provider.max_completion_tokens: 262140
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing increased and output token limit quadrupled. Direct impact on engineering budgets and application design.
signalsprompt pricing increased ~6.7% · completion pricing increased ~35% · max_completion_tokens increased 4x (65536 → 262140) · pricing and limits are explicit commitments affecting cost and capability
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGLIMITSMODEL OpenRouter model catalogue · 2026-09-22 16:27 UTC

OpenAI gpt-oss-20b pricing cut 40% prompt, 31% completion; output capped at 32K tokens

DOCheck your application logs for any gpt-oss-20b completions exceeding 32,768 tokens in the past 30 days.
WHAT CHANGEDOpenRouter's gpt-oss-20b model pricing dropped from $0.00000003 to $0.00000018 per prompt token (40% reduction) and $0.00000013 to $0.00000009 per completion token (31% reduction). Maximum completion tokens reduced from 117,964 to 32,768, effective now.
WHY IT MATTERSIf you generate responses longer than 32K tokens, calls will now truncate or fail. Cost forecasts based on the old pricing are overstated by roughly a third.
WHAT TO DOCheck your application logs for any gpt-oss-20b completions exceeding 32,768 tokens in the past 30 days. If found, either chunk your outputs or switch models. Update your cost model with the new per-token rates.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: openai/gpt-oss-20b
 name: OpenAI: gpt-oss-20b
-pricing.prompt: 0.00000003
-pricing.completion: 0.00000013
+pricing.prompt: 0.000000018
+pricing.completion: 0.00000009
 context_length: 131072
-top_provider.max_completion_tokens: 117964
+top_provider.max_completion_tokens: 32768
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing dropped significantly and max output tokens cut by 72%, affecting budget and use-case viability.
signalspricing.prompt decreased 40% · pricing.completion decreased 31% · max_completion_tokens reduced by 72% (117964 → 32768) · model capability constraint change
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELLIMITS OpenRouter model catalogue · 2026-09-22 16:27 UTC

AionLabs Aion-3.0-Mini context window expanded to 1,048,576 tokens

DOIdentify pipelines currently splitting inputs for this model and benchmark end-to-end cost and latency with single-pass processing at the new limit.
WHAT CHANGEDThe context_length for AionLabs: Aion-3.0-Mini increased from 131,072 tokens to 1,048,576 tokens (8x expansion). Pricing remains $0.0000007 per prompt token and $0.0000014 per completion token.
WHY IT MATTERSApplications currently chunking or truncating inputs to fit the 131K limit can now process 8x more content in a single request, eliminating multi-pass workflows and reducing latency. This changes the cost-performance calculation for document analysis, long-context retrieval, and conversation history retention.
WHAT TO DOIdentify pipelines currently splitting inputs for this model and benchmark end-to-end cost and latency with single-pass processing at the new limit. Update any hardcoded context_length checks from 131072 to 1048576.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 name: AionLabs: Aion-3.0-Mini
 pricing.prompt: 0.0000007
 pricing.completion: 0.0000014
-context_length: 131072
+context_length: 1048576
 top_provider.max_completion_tokens: 32768
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningContext window expansion is a material capability change affecting what workloads the model can handle.
signalscontext_length increased 8x (131072 → 1048576 tokens) · model capability change affects application architecture decisions
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELLIMITS OpenRouter model catalogue · 2026-09-22 16:27 UTC

AionLabs Aion-3.0 context window expanded from 131K to 1M tokens

DOIdentify code paths that chunk documents or conversations specifically for Aion-3.0's previous 131K limit, and test whether removing that chunking logic improves perform…
WHAT CHANGEDThe context_length for AionLabs: Aion-3.0 increased from 131,072 tokens to 1,048,576 tokens (8x expansion). Pricing remains unchanged at $0.000003 per prompt token and $0.000006 per completion token.
WHY IT MATTERSApplications currently chunking or summarizing documents to fit within 131K tokens can now process 8x larger inputs in a single call, reducing complexity and potentially lowering total costs by eliminating multi-pass operations.
WHAT TO DOIdentify code paths that chunk documents or conversations specifically for Aion-3.0's previous 131K limit, and test whether removing that chunking logic improves performance or reduces API calls.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 name: AionLabs: Aion-3.0
 pricing.prompt: 0.000003
 pricing.completion: 0.000006
-context_length: 131072
+context_length: 1048576
 top_provider.max_completion_tokens: 32768
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningContext window expansion is a material capability change affecting prompt design and use-case feasibility.
signalscontext_length increased 8x (131072 → 1048576 tokens) · model capability change affects application design decisions
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING MODELLIMITS OpenRouter model catalogue · 2026-09-22 16:27 UTC

AionLabs Aion-2.0 context window expanded from 131K to 1M tokens

DOIdentify codepaths that chunk inputs for this model and evaluate whether removing chunking logic would reduce latency or cost.
WHAT CHANGEDThe context_length for AionLabs: Aion-2.0 increased from 131,072 tokens to 1,048,576 tokens (8x expansion). Pricing remains unchanged at $0.0000008 per prompt token and $0.0000016 per completion token.
WHY IT MATTERSApplications currently chunking or truncating inputs to fit the 131K limit can now process 8x more context in a single call, potentially eliminating multi-pass strategies and reducing total API calls for large document processing.
WHAT TO DOIdentify codepaths that chunk inputs for this model and evaluate whether removing chunking logic would reduce latency or cost. Test with representative large inputs to confirm the expanded window works as expected.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 name: AionLabs: Aion-2.0
 pricing.prompt: 0.0000008
 pricing.completion: 0.0000016
-context_length: 131072
+context_length: 1048576
 top_provider.max_completion_tokens: 32768
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningContext window expansion is a material capability change affecting how engineers can use this model.
signalscontext_length increased 8x (131072 → 1048576 tokens) · model capability change affects application design decisions
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING LIMITSAPIMODEL OpenRouter model catalogue · 2026-09-22 11:06 UTC

OpenRouter max completion tokens increased from 102400 to 943718

DOGrep your codebase for max_tokens or max_completion_tokens parameters in OpenRouter calls.
WHAT CHANGEDThe top_provider.max_completion_tokens limit increased from 102,400 to 943,718 tokens (9.2x increase). Context length remains 1,310,720 tokens. Pricing unchanged at $0.000000075 per prompt token and $0.00000025 per completion token.
WHY IT MATTERSIf you currently cap completion requests at the old 102,400 limit in your code, you are leaving 841,318 tokens of available capacity unused. Conversely, if you request completions without explicit limits, responses could now be 9x larger than before, increasing latency and cost per request by the same factor.
WHAT TO DOGrep your codebase for max_tokens or max_completion_tokens parameters in OpenRouter calls. If you set explicit caps, decide whether to raise them. If you rely on defaults, add explicit caps now to prevent unexpectedly large responses.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 pricing.prompt: 0.000000075
 pricing.completion: 0.00000025
 context_length: 1310720
-top_provider.max_completion_tokens: 102400
+top_provider.max_completion_tokens: 943718
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.92, needed 0.5
triage reasoningCompletion token limit nearly 10x higher; engineering teams must re-evaluate request sizing and costs.
signalsmax_completion_tokens increased 9.2x (102400 → 943718) · limit change on production model · affects token budgeting and request planning
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING PRICINGMODEL OpenRouter model catalogue · 2026-09-22 11:06 UTC

DeepSeek V4 Flash completion tokens increased 25%

DOPull your last 30 days of completion token usage for this model ID from OpenRouter logs or your own metrics, multiply by the new rate, and update your cost forecast.
WHAT CHANGEDOpenRouter raised the completion token price for deepseek/deepseek-v4-flash-latest from $0.0000008 to $0.000001 per token, a 25% increase. Prompt pricing ($0.00000003) and context length (1.31M tokens) remain unchanged.
WHY IT MATTERSIf you generate high volumes of completion tokens with this model, your monthly API costs will increase by 25% at current usage levels. A workload generating 1 billion completion tokens per month now costs $1,000 instead of $800.
WHAT TO DOPull your last 30 days of completion token usage for this model ID from OpenRouter logs or your own metrics, multiply by the new rate, and update your cost forecast. If the increase is material, evaluate whether switching to the non-latest pinned version or a different model makes sense.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-v4-flash-latest
 name: DeepSeek: DeepSeek V4 Flash Latest
 pricing.prompt: 0.00000003
-pricing.completion: 0.0000008
+pricing.completion: 0.000001
 context_length: 1310720
 top_provider.max_completion_tokens: 943718
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.95, needed 0.5
triage reasoningCompletion pricing for DeepSeek V4 Flash increased 25%, affecting cost calculations for production workloads.
signalscompletion token price increased · price change: 0.0000008 → 0.000001 (25% increase)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
ACTION NEEDED PRICINGMODEL OpenRouter model catalogue · 2026-09-22 11:06 UTC

DeepSeek Pro Latest pricing increased 12% prompt, 72% completion

DOPull last 30 days of deepseek-pro-latest usage from OpenRouter logs, calculate new cost at current volume, and update your monthly forecast.
WHAT CHANGEDOpenRouter's deepseek-pro-latest model pricing changed: prompt tokens from $0.000000558624 to $0.000000624 per token (+11.7%), completion tokens from $0.000001675872 to $0.00000288 per token (+71.6%). Max completion tokens increased from 384,000 to 943,718.
WHY IT MATTERSCompletion token costs rose 72%, which compounds on long-form generation workloads. A 10,000-token completion that cost $0.0168 now costs $0.0288, a $0.012 increase per call. High-volume applications will see material budget impact.
WHAT TO DOPull last 30 days of deepseek-pro-latest usage from OpenRouter logs, calculate new cost at current volume, and update your monthly forecast. If completion length averages above 5,000 tokens, evaluate whether the increased max_completion_tokens justifies the 72% cost increase or whether to switch models.
CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~deepseek/deepseek-pro-latest
 name: DeepSeek: DeepSeek Pro Latest
-pricing.prompt: 0.000000558624
-pricing.completion: 0.000001675872
+pricing.prompt: 0.000000624
+pricing.completion: 0.00000288
 context_length: 1048576
-top_provider.max_completion_tokens: 384000
+top_provider.max_completion_tokens: 943718
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing per token increased significantly; max output tokens nearly doubled.
signalspricing.prompt increased ~11.7% · pricing.completion increased ~71.6% · max_completion_tokens increased ~145% (384000 → 943718)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)
WORTH KNOWING LIMITSMODELAPI OpenRouter model catalogue · 2026-09-22 11:06 UTC

OpenRouter max_completion_tokens increased from 32768 to 235929

DOIf you have workflows that chunk outputs to stay under 32768 tokens, test whether single-pass generation at higher token counts improves quality or reduces latency for y…
WHAT CHANGEDThe top_provider.max_completion_tokens value increased from 32768 to 235929 tokens. Context length remains 262144 tokens. Pricing unchanged at $0.00000015 per prompt token and $0.0000012 per completion token.
WHY IT MATTERSApplications currently capped at 32768 output tokens can now request up to 235929 tokens per completion, enabling longer-form generation without chunking. Existing requests under the old limit continue to work unchanged.
WHAT TO DOIf you have workflows that chunk outputs to stay under 32768 tokens, test whether single-pass generation at higher token counts improves quality or reduces latency for your use case.
CONFIDENCE 85% Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 pricing.prompt: 0.00000015
 pricing.completion: 0.0000012
 context_length: 262144
-top_provider.max_completion_tokens: 32768
+top_provider.max_completion_tokens: 235929
HOW IT WAS JUDGED
triage verdictMATERIAL at 0.92, needed 0.5
triage reasoningMaximum completion token limit changed substantially, affecting application design and output capacity.
signalsmax_completion_tokens decreased significantly · 32768 → 235929 represents constraint change · affects output capacity for dependent applications
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)