Claude Opus 5
Claude Opus 5 surpassed Mistral Large 4 in the blind human evaluation. Kimi‑K3 outperforms Claude Opus 5.5 on the ThinkingBox benchmark. Claude Opus 5.5 outperforms Claude Opus 5 on every-attempt average. Chirp Pong is built using Claude Opus 5.5 via Claude Code. Claude traced the seeds to the ordinary random‑number generator lrand48. Deepseek V4.1‑Flash‑1017 outperforms Claude Opus 5. Claude is version Claude Opus 5.5. ArtCraft was built with Claude Opus 5.5. Anthropic released Claude Opus 5.5 on September 22, 2026. Claude Opus 5.5 is compared with Claude Fable 5.1. Muse Wonder Full beats Claude Opus 5.5. Artificial Analysis ranked Claude Opus 5.5 first. Personal Superintelligence Index ranked Claude Opus 5.5. Gemini 4 Argon trails Claude Opus 5.5 in independent benchmarks. Gemini 4 Argon model is priced at half the price of Claude Opus 5.5. On 2026‑12‑01 the Gemini 4 Argon model’s price will be equal to that of Claude Opus 5.5. Claude Duet is built around Claude Opus 5.5. GPT‑6.1 Terra has a lower Intelligence Index score (48) than Claude Opus 5.5’s score (58) Anthropic provides launch notes for Claude Opus 5.5 Gemini 4 Carbon outperformed Claude Opus 5.5 on ExploitGym tests Gemini 4 Carbon has a median task cost of $1.80, which is 72% lower than Claude Opus 5.5 Gemini 4 Carbon has 70.36% accuracy in the Vals Index v2.2, higher than Claude Opus 5.5 70.18% accuracy Claude Opus 5.5 tops the Artificial Analysis' Coding Agent Index Claude Opus 5.5 currently sits atop the Epoch Capabilities Index· Entity summary generated by Graphiti
Relation timeline
BY VALID-FROM DATE- Gemini 4 ArgonWILL BE EQUAL PRICE ASClaude Opus 5
On 2026‑12‑01 the Gemini 4 Argon model’s price will be equal to that of Claude Opus 5.5
- Claude Opus 5SITS AT TOP OFEpoch Capabilities Index
Claude Opus 5.5 currently sits atop the Epoch Capabilities Index
- Gemini 4 CarbonLOWER TASK COST THANClaude Opus 5
Gemini 4 Carbon has a median task cost of $1.80, which is 72% lower than Claude Opus 5.5
- Gemini 4 CarbonHIGHER ACCURACY THANClaude Opus 5
Gemini 4 Carbon has 70.36% accuracy in the Vals Index v2.2, higher than Claude Opus 5.5 70.18% accuracy
- Gemini 4 CarbonOUTPERFORMEDClaude Opus 5
Gemini 4 Carbon outperformed Claude Opus 5.5 on ExploitGym tests
- AnthropicPROVIDESClaude Opus 5
Anthropic provides launch notes for Claude Opus 5.5
- GPT-6.1 TerraHAS LESS SCORE THANClaude Opus 5
GPT‑6.1 Terra has a lower Intelligence Index score (48) than Claude Opus 5.5’s score (58)
- GPT-6.1 TerraCLOSE TO SCORE WITHIN FIVE POINTS TOClaude Opus 5
GPT-6.1 Terra comes within five points of Claude Opus 5.5 at less than a fifth of the estimated cost on Terminal‑Bench‑Science 0.1
- GPT-6.1 TerraCOST PER TASK IS NINE TIMES LESS THANClaude Opus 5
GPT-6.1 Terra costs roughly nine times less per task than Claude Opus 5.5
- GPT-6.1 TerraTOKEN EFFICIENT MULTIPLIED OVERClaude Opus 5
GPT-6.1 Terra is more than three times token‑efficient compared to Claude Opus 5.5 at maximum effort
- Claude Opus 5BEATSClaude Opus 5
Claude Opus 5.5 beats its predecessor Claude Opus 5 in 19 out of 21 comparisons
- AnthropicPRODUCEDClaude Opus 5
Anthropic produced the model Claude Opus 5
- AnthropicPRODUCEDClaude Opus 5
Anthropic produced the model Claude Opus 5.5
- webframeEVALUATED BYClaude Opus 5
Claude Opus 5.5 evaluated webframe
- Claude Opus 5SCOREDProgramBench
Claude Opus 5.5 scored 18.5% on ProgramBench.
- Claude Opus 5WORKS FOROpenAI
Eric Provencher is an OpenAI developer.
- Claude Opus 5HAS HARNESS WITH LEAD OVERClaude Code
Zhipu AI has a harness with a 2.4-point lead over Claude Code harness
- Kimi K3.1 ProHALF OUTPUT PRICE OFClaude Opus 5
Kimi K3.1 Pro has an output price that is half that of Claude Opus 5.5
- Kimi K3.1 ProIS BEHINDClaude Opus 5
Kimi K3.1 Pro is behind Claude Opus 5.5 in Artificial Analysis scoring
- Kimi K3.1 ProAVAILABLE THROUGHClaude Opus 5
Kimi K3.1 Pro is available through Kimi.ai
- T4T5USEDClaude Opus 5
T4T5 USED Claude Opus 5.5 to reconstruct Satoshi’s computer setup
- Deepseek V4.1-Flash-1017OUTPERFORMSClaude Opus 5
Deepseek V4.1-Flash-1017 outperforms Claude Opus 5
- ClaudeTRACED SEEDS TOClaude Opus 5
Claude traced the seeds to the ordinary random‑number generator lrand48.
- Chirp PongBUILT WITHClaude Opus 5
Chirp Pong is built using Claude Opus 5.5 via Claude Code.
- ArtCraftBUILT WITHClaude Opus 5
ArtCraft was built with Claude Opus 5.5
- Nathan FielderBUILT WITHClaude Opus 5
ArtCraft was built with Claude Opus 5.5
- ClaudeVERSIONClaude Opus 5
Claude is version Claude Opus 5.5
- Gemini 4 ArgonHALF PRICE OFClaude Opus 5
Gemini 4 Argon model is priced at half the price of Claude Opus 5.5
- Gemini 4 ArgonTRAILSClaude Opus 5
Gemini 4 Argon trails Claude Opus 5.5 in independent benchmarks
- Personal Superintelligence IndexRANKEDClaude Opus 5
Personal Superintelligence Index ranked Claude Opus 5.5
- Claude Opus 5COMPARED WITHClaude Fable 5.1
Claude Opus 5.5 is compared with Claude Fable 5.1
- Muse Wonder FullBEATSClaude Opus 5
Muse Wonder Full beats Claude Opus 5.5
- Artificial AnalysisRANKEDClaude Opus 5
Artificial Analysis ranked Claude Opus 5.5 first
- Brandon ThomasUTILIZESClaude Opus 5
Brandon Thomas utilizes Claude Opus 5.5 to create ArtCraft.
- Claude Opus 5SURPASSEDMistral Large 4
Claude Opus 5 surpassed Mistral Large 4 in the blind human evaluation
- Kimi K3OUTPERFORMSClaude Opus 5
Kimi-K3 outperforms Claude Opus 5.5 on the ThinkingBox benchmark
- Claude Opus 5OUTPERFORMSClaude Opus 5
Claude Opus 5.5 outperforms Claude Opus 5 on every-attempt average
- AnthropicRELEASEDClaude Opus 5
Anthropic released Claude Opus 5.5 on September 22, 2026
- Claude Opus 5TOPS INArtificial Analysis' Coding Agent Index
Claude Opus 5.5 tops the Artificial Analysis' Coding Agent Index
- Claude DuetBASED ONClaude Opus 5
Claude Duet is built around Claude Opus 5.5
Relations are facts Graphiti extracted from reports, not verbatim quotes. An end date means later reporting showed the relation no longer holds.
Events 12
- Paris-based Mistral releases "Le Chonk," a strong coding model that still trails China overall1 cited report · 2026-10-29
- Anthropic’s business plan: Build God or die trying1 cited report · 2026-10-29
- Google Deepmind releases Gemini 4 Carbon just two weeks after Argon with gains across the board3 cited reports · 2026-10-29
- How Anthropic ships the #1 LLM in the World safely: Separating the Brain from the Hands4 cited reports · 2026-10-28
- Why OpenAI Is Selling GPT-6.1 Terra on Price, Not Benchmarks2 cited reports · 2026-10-27
- Researchers find six patterns that make AI text so exhausting to read1 cited report · 2026-10-27
- webframe: evaluated as a scraper and a sandbox1 cited report · 2026-10-27
- Import AI 478: Math for alignment; ProgramBench; and Mirendil shows what RSI means1 cited report · 2026-10-26
- Anthropic says one AI agent beats many subagents now that models can handle complexity alone1 cited report · 2026-10-25
- Fancy harnesses make coding agents more expensive but not better, study finds1 cited report · 2026-10-24
- Introducing Kimi K3.1 Pro1 cited report · 2026-10-24
- Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs1 cited report · 2026-10-17
Source reports 29
- Paris-based Mistral releases "Le Chonk," a strong coding model that still trails China overallThe Decoder · 2026-10-29
- Anthropic’s business plan: Build God or die tryingThe Verge · AI · 2026-10-29
- Google Deepmind releases Gemini 4 Carbon just two weeks after Argon with gains across the boardThe Decoder · 2026-10-29
- Google announces Gemini 4 Carbon, its latest frontier AI model that you can't useArs Technica · AI · 2026-10-29
- Gemini 4 Carbon Beats Claude and GPT-6: Why Is Google Holding It Back?BeInCrypto · 2026-10-29
- How Anthropic ships the #1 LLM in the World safely: Separating the Brain from the HandsLatent Space · 2026-10-28
- Why OpenAI Is Selling GPT-6.1 Terra on Price, Not BenchmarksBeInCrypto · 2026-10-27
- GPT-6.1 Terra halves OpenAI's API prices and still gains ground on its larger modelsThe Decoder · 2026-10-27
- Researchers find six patterns that make AI text so exhausting to readThe Decoder · 2026-10-27
- webframe: evaluated as a scraper and a sandboxSimon Willison · 2026-10-27
- Import AI 478: Math for alignment; ProgramBench; and Mirendil shows what RSI meansImport AI · 2026-10-26
- Anthropic says one AI agent beats many subagents now that models can handle complexity aloneThe Decoder · 2026-10-25
- Fancy harnesses make coding agents more expensive but not better, study findsThe Decoder · 2026-10-24
- Introducing Kimi K3.1 ProSimon Willison · 2026-10-24
- Chinese AI model Kimi K3.1 Pro is closing in on Anthropic's Claude Opus 5.5The Decoder · 2026-10-24
- Claude AI Cracks Satoshi’s 17-Year Bitcoin MysteryBeInCrypto · 2026-10-23
- Anthropic takes on Grok Bot, Meta's Muse, and OpenAI's Dots with its always-on agent "Claude Duet"The Decoder · 2026-10-22
- CZ Reacts to Anthropic’s New Claude Duet AI Agent: ‘Now, It Controls Us’BeInCrypto · 2026-10-22
- Anthropic takes on Muse with Claude Duet, its new AI agentThe Verge · AI · 2026-10-22
- Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costsThe Decoder · 2026-10-17
- ‘Wow, it’s a real discovery’: Claude cracks a 25-year NSA number mysteryCoinDesk · 2026-10-17
- Chirp PongSimon Willison · 2026-10-16
- Developer Who Cloned Photoshop With AI Just Did the Same to Microsoft OfficeDecrypt · 2026-10-15
- llm-gemini 0.35Simon Willison · 2026-10-14
- Gemini 4 Argon is now available worldwide as Google's most capable frontier modelThe Decoder · 2026-10-14
- Meta's New Muse Wonder Finally Puts Zuckerberg in the Frontier AI RaceDecrypt · 2026-10-13
- From GTA V on iPhone to Free Photoshop: This AI Tool Is Upending SoftwareDecrypt · 2026-10-13
- Introducing Mistral Large 4Mistral AI · first-party · 2026-10-06
- The Agent Said It Was Done. The Database Disagreed.Hugging Face Blog · first-party · 2026-10-04