Data observatory

Capability.
In context.

General benchmarks help identify candidate models. Our operational work asks whether those capabilities survive the instructions, sources, handoffs and review requirements of a particular task.

External measurements

Source snapshot retrieved 2026-10-16T15:15:44.060Z. These are source-reported results, not MEKreflect reruns. Coverage is limited to the records available in this release. Missing scores are unknown, never zero.

Sources, reuse and refresh status
Leading model groups in the imported Epoch Capabilities Index snapshotTwelve highest source scores, with source confidence bounds when supplied. External measurements by Epoch AI, not MEKreflect evaluations.03672108144180Claude Opus 5.5166.8GPT-6 Astra166.4GPT-6.1 Sol166.3Claude Fable 5.1164.7Gemini 4 Argon164.6Claude Sonnet 5.5164.4Claude Opus 5162.6GPT-6 Sol162.6Claude Fable 5162.1GPT-5.5 Pro162.0GPT-5.6 Sol161.7GPT-5.6 Terra159.4Epoch Capabilities Index · source scale
Epoch AI data, CC BY 4.0. MEKreflect selected and visualised twelve model groups. Lines around dots show source-supplied confidence bounds. This is a dated snapshot, not a live ranking. Original method and current results.

277 records

Source-reported model records
ModelOrganisationMeasureValueSource
Claude Opus 5.5AnthropicEpoch Capabilities Index166.82Source
GPT-6 AstraOpenAIEpoch Capabilities Index166.41Source
GPT-6.1 SolOpenAIEpoch Capabilities Index166.32Source
Claude Fable 5.1AnthropicEpoch Capabilities Index164.65Source
Gemini 4 ArgonGoogle DeepMindEpoch Capabilities Index164.56Source
Claude Sonnet 5.5AnthropicEpoch Capabilities Index164.45Source
Claude Opus 5AnthropicEpoch Capabilities Index162.6Source
GPT-6 SolOpenAIEpoch Capabilities Index162.55Source
Claude Fable 5AnthropicEpoch Capabilities Index162.13Source
GPT-5.5 ProOpenAIEpoch Capabilities Index161.97Source
GPT-5.6 SolOpenAIEpoch Capabilities Index161.73Source
GPT-5.6 TerraOpenAIEpoch Capabilities Index159.43Source
GPT-5.4 ProOpenAIEpoch Capabilities Index159Source
Claude Haiku 5.5AnthropicEpoch Capabilities Index158.98Source
GPT-5.5OpenAIEpoch Capabilities Index158.95Source
Claude Opus 4.8AnthropicEpoch Capabilities Index158.26Source
Kimi K3MoonshotEpoch Capabilities Index157.42Source
Gemini 3.7 FlashGoogle DeepMindEpoch Capabilities Index157.22Source
Gemini 3.8 FlashGoogle DeepMindEpoch Capabilities Index157.2Source
Qwen 3.8 MaxAlibabaEpoch Capabilities Index156.88Source
GPT-5.3 CodexOpenAIEpoch Capabilities Index156.76Source
Muse Spark 1.3Meta AIEpoch Capabilities Index156.76Source
GPT-5.4OpenAIEpoch Capabilities Index156.72Source
Claude Sonnet 5AnthropicEpoch Capabilities Index156.36Source
GPT-5.6 LunaOpenAIEpoch Capabilities Index156.33Source
Claude Opus 4.7AnthropicEpoch Capabilities Index156.32Source
GPT-6 LunaOpenAIEpoch Capabilities Index156.31Source
Grok 4.6xAIEpoch Capabilities Index156.31Source
GPT-5.2 ProOpenAIEpoch Capabilities Index156.03Source
GLM-5.3Z.ai (Zhipu AI)Epoch Capabilities Index155.91Source
DeepSeek V4.1 FlashDeepSeekEpoch Capabilities Index155.57Source
DeepSeek V4 Pro 0813DeepSeekEpoch Capabilities Index155.29Source
Claude Opus 4.6AnthropicEpoch Capabilities Index155.27Source
Qwen3.8 Max (0902)AlibabaEpoch Capabilities Index154.91Source
Muse Spark 1.2Meta AIEpoch Capabilities Index154.9Source
Gemini 3.1 ProGoogle DeepMindEpoch Capabilities Index154.67Source
DeepSeek V4 Flash 0731DeepSeekEpoch Capabilities Index154.49Source
Gemini 3.5 FlashGoogle DeepMindEpoch Capabilities Index154.38Source
Grok 4.7xAIEpoch Capabilities Index154.37Source
Gemini 3.6 FlashGoogle DeepMindEpoch Capabilities Index154.24Source
Muse Spark 1.1Meta AIEpoch Capabilities Index154Source
Grok 4.5xAIEpoch Capabilities Index153.93Source
Qwen3.7-MaxAlibabaEpoch Capabilities Index153.83Source
GPT-5.2OpenAIEpoch Capabilities Index153.46Source
GLM-5.3-FlashZ.ai (Zhipu AI)Epoch Capabilities Index153.23Source
Gemini 3 ProGoogle DeepMindEpoch Capabilities Index152.9Source
Claude Sonnet 4.6AnthropicEpoch Capabilities Index152.39Source
Muse SparkMeta AIEpoch Capabilities Index152.13Source
Grok 4.20xAIEpoch Capabilities Index152.08Source
GLM-5.2Z.ai (Zhipu AI)Epoch Capabilities Index151.86Source
Gemini 3 FlashGoogle DeepMindEpoch Capabilities Index151.8Source
Qwen 3.8 27BAlibabaEpoch Capabilities Index151.55Source
Kimi K2.6MoonshotEpoch Capabilities Index150.92Source
GPT-5 ProOpenAIEpoch Capabilities Index150.74Source
Inkling-SmallThinking MachinesEpoch Capabilities Index150.18Source
Claude Opus 4.5AnthropicEpoch Capabilities Index150.17Source
GPT-5OpenAIEpoch Capabilities Index150Source
GLM-5.1Z.ai (Zhipu AI)Epoch Capabilities Index149.99Source
Kimi K2.7 CodeMoonshotEpoch Capabilities Index149.87Source
GPT-5.1OpenAIEpoch Capabilities Index149.76Source
Qwen 3.6 Max (Preview)AlibabaEpoch Capabilities Index149.24Source
DeepSeek-V4-ProDeepSeekEpoch Capabilities Index149.19Source
Grok 4.3 BetaxAIEpoch Capabilities Index148.95Source
GPT-5.4 MiniOpenAIEpoch Capabilities Index148.77Source
InklingThinking MachinesEpoch Capabilities Index148.59Source
Kimi K2.5MoonshotEpoch Capabilities Index147.98Source
Qwen 3.6 PlusAlibabaEpoch Capabilities Index147.49Source
o3-proOpenAIEpoch Capabilities Index147.37Source
Qwen3.7-PlusAlibabaEpoch Capabilities Index147.3Source
MiniMax-M3MiniMaxEpoch Capabilities Index147.13Source
Mistral Large 4Mistral AIEpoch Capabilities Index146.9Source
o3OpenAIEpoch Capabilities Index146.87Source
Claude Sonnet 4.5AnthropicEpoch Capabilities Index146.83Source
MiniMax-M2.5MiniMaxEpoch Capabilities Index146.67Source
Qwen3.5 397B-A17BAlibabaEpoch Capabilities Index146.66Source
Qwen3.6 27BAlibabaEpoch Capabilities Index146.48Source
Grok 4xAIEpoch Capabilities Index146.43Source
Qwen 3.5 Plus (hosted 397B-A17B)AlibabaEpoch Capabilities Index146.32Source
GPT-5.4 NanoOpenAIEpoch Capabilities Index146.12Source
DeepSeek-V3.2DeepSeekEpoch Capabilities Index146.11Source
DeepSeek-V4-FlashDeepSeekEpoch Capabilities Index146.05Source
Kimi K2 ThinkingMoonshotEpoch Capabilities Index145.96Source
MiniMax-M2.7MiniMaxEpoch Capabilities Index145.88Source
GLM-5Z.ai (Zhipu AI)Epoch Capabilities Index145.78Source
Nemotron 3 UltraNvidiaEpoch Capabilities Index145.77Source
o4-miniOpenAIEpoch Capabilities Index145.6Source
GPT-5 miniOpenAIEpoch Capabilities Index145.43Source
Gemini 2.5 Pro (Jun 2025)Google DeepMindEpoch Capabilities Index145.31Source
Gemini 3.5 Flash-LiteGoogle DeepMindEpoch Capabilities Index145.15Source
DeepSeek-V3.2-ExpDeepSeekEpoch Capabilities Index144.97Source
Qwen3.7 FlashAlibabaEpoch Capabilities Index144.64Source
Gemini 3.1 Flash-LiteGoogleEpoch Capabilities Index144.47Source
Grok 4 FastxAIEpoch Capabilities Index144.19Source
Gemini 2.5 Pro (Mar 2025)Google DeepMindEpoch Capabilities Index144.12Source
Qwen 3.6 35B-A3BAlibabaEpoch Capabilities Index144.11Source
Claude Opus 4.1AnthropicEpoch Capabilities Index144.09Source
Qwen3-235B-A22B-Thinking (Jul 2025)AlibabaEpoch Capabilities Index143.81Source
GLM-4.7Z.ai (Zhipu AI)Epoch Capabilities Index143.79Source
Qwen 3.5 Flash (hosted 35B-A3B)AlibabaEpoch Capabilities Index143.77Source
Qwen 3.6 FlashAlibabaEpoch Capabilities Index143.05Source
Gemini 2.5 Flash (Sep 2025)Google DeepMindEpoch Capabilities Index143Source
Gemma 4 31B ITGoogle DeepMindEpoch Capabilities Index142.86Source
Claude Opus 4AnthropicEpoch Capabilities Index142.65Source
Qwen3.5-35B-A3BAlibabaEpoch Capabilities Index142.48Source
GPT-5.5 InstantOpenAIEpoch Capabilities Index142.45Source
Gemini 2.5 Pro (May 2025)Google DeepMindEpoch Capabilities Index142.44Source
Claude Haiku 4.5AnthropicEpoch Capabilities Index142.4Source
Qwen3-MaxAlibabaEpoch Capabilities Index142.34Source
o1OpenAIEpoch Capabilities Index141.84Source
Gemma 4 26B A4BGoogle DeepMindEpoch Capabilities Index141.81Source
Claude Sonnet 4AnthropicEpoch Capabilities Index141.67Source
Gemini 2.5 Flash (May 2025)Google DeepMindEpoch Capabilities Index141.51Source
Mistral Medium 3.5Mistral AIEpoch Capabilities Index141.47Source
DeepSeek-R1 (May 2025)DeepSeekEpoch Capabilities Index141.27Source
Claude 3.7 SonnetAnthropicEpoch Capabilities Index141.14Source
Gemini 2.5 Flash (Jun 2025)Google DeepMindEpoch Capabilities Index140.77Source
o3-miniOpenAIEpoch Capabilities Index140.33Source
Grok-3 minixAIEpoch Capabilities Index140.33Source
Kimi K2 (Jul 2025)MoonshotEpoch Capabilities Index140.12Source
gpt-oss-120bOpenAIEpoch Capabilities Index140Source
Gemini 2.5 Flash (Apr 2025)Google DeepMindEpoch Capabilities Index139.94Source
DeepSeek-V3.1DeepSeekEpoch Capabilities Index139.91Source
Qwen3-30B-A3B-Thinking (Jul 2025)AlibabaEpoch Capabilities Index139.61Source
Qwen3.5-9BAlibabaEpoch Capabilities Index139.43Source
Qwen3-235B-A22BAlibabaEpoch Capabilities Index139.35Source
GPT-5 nanoOpenAIEpoch Capabilities Index139.33Source
DeepSeek-R1DeepSeekEpoch Capabilities Index138.96Source
Qwen3-235B-A22B-Instruct (Jul 2025)AlibabaEpoch Capabilities Index138.92Source
Qwen3-32BAlibabaEpoch Capabilities Index138.49Source
Grok 3xAIEpoch Capabilities Index138.32Source
Qwen3-14BAlibabaEpoch Capabilities Index138.22Source
gpt-oss-20bOpenAIEpoch Capabilities Index137.81Source
QwQ-32BAlibabaEpoch Capabilities Index137.58Source
DeepSeek-R1-Distill-Qwen-32BDeepSeekEpoch Capabilities Index137.42Source
Qwen3-30B-A3B-Instruct (Jul 2025)AlibabaEpoch Capabilities Index137.42Source
GPT-4.1OpenAIEpoch Capabilities Index136.77Source
GPT-4.5OpenAIEpoch Capabilities Index136.73Source
Qwen3-30B-A3BAlibabaEpoch Capabilities Index136.19Source
Qwen3-8BAlibabaEpoch Capabilities Index136.17Source
DeepSeek-V3 (Mar 2025)DeepSeekEpoch Capabilities Index135.93Source
o1-miniOpenAIEpoch Capabilities Index135.82Source
DeepSeek-R1-Distill-Qwen-14BDeepSeekEpoch Capabilities Index135.44Source
Gemini 2.0 Flash Thinking (Jan 2025)Google DeepMind,GoogleEpoch Capabilities Index135.36Source
Gemini 2.0 ProGoogle DeepMindEpoch Capabilities Index135.05Source
GPT-4.1 miniOpenAIEpoch Capabilities Index135Source
o1-previewOpenAIEpoch Capabilities Index134.78Source
Gemini 2.0 Flash (Dec 2024)Google DeepMind,GoogleEpoch Capabilities Index134.69Source
Gemini 2.0 Flash (Feb 2025)Google DeepMind,GoogleEpoch Capabilities Index134.68Source
Mistral Medium 3Mistral AIEpoch Capabilities Index134.06Source
Gemini 2.5 Flash-Lite (Jun 2025)Google DeepMindEpoch Capabilities Index133.95Source
Claude 3.5 Sonnet (October 2024)AnthropicEpoch Capabilities Index133.53Source
Magistral Small 1.0Mistral AIEpoch Capabilities Index133.2Source
Qwen2.5-MaxAlibabaEpoch Capabilities Index132.53Source
DeepSeek-V3DeepSeekEpoch Capabilities Index132.33Source
Llama 4 MaverickMeta AIEpoch Capabilities Index132.2Source
Mistral Small 3.2Mistral AIEpoch Capabilities Index131.75Source
Gemini 1.5 Pro (Sept 2024)Google DeepMindEpoch Capabilities Index131.73Source
Magistral Small 1.2Mistral AIEpoch Capabilities Index131.43Source
Grok-2 (Dec 2024)xAIEpoch Capabilities Index130.49Source
Phi-4Microsoft ResearchEpoch Capabilities Index130.43Source
Gemma 3 27BGoogle DeepMindEpoch Capabilities Index130.05Source
Claude 3.5 SonnetAnthropicEpoch Capabilities Index130Source
Llama 4 ScoutMeta AIEpoch Capabilities Index129.65Source
GPT-4.1 nanoOpenAIEpoch Capabilities Index129.63Source
Gemini 1.5 Flash (Sep 2024)Google DeepMindEpoch Capabilities Index129.37Source
GPT-4o (May 2024)OpenAIEpoch Capabilities Index129.05Source
Qwen2.5-72BAlibabaEpoch Capabilities Index129Source
GPT-4o (Aug 2024)OpenAIEpoch Capabilities Index128.84Source
Llama 3.1-405BMeta AIEpoch Capabilities Index128.76Source
GPT-4o (Nov 2024)OpenAIEpoch Capabilities Index128.73Source
Qwen2.5-32BAlibabaEpoch Capabilities Index128.53Source
Mistral Large 2 (Nov 2024)Mistral AIEpoch Capabilities Index128.53Source
Mistral Large 2 (Jul 2024)Mistral AIEpoch Capabilities Index127.55Source
Mistral Small 3.1Mistral AIEpoch Capabilities Index127.49Source
Llama 3.3 70BMeta AIEpoch Capabilities Index127.34Source
GPT-4 Turbo (Apr 2024)OpenAIEpoch Capabilities Index127.25Source
Claude 3.5 HaikuAnthropicEpoch Capabilities Index127.17Source
Mistral Small 3Mistral AIEpoch Capabilities Index127.08Source
Gemini 1.5 Pro (May 2024)Google DeepMindEpoch Capabilities Index126.92Source
Claude 3 OpusAnthropicEpoch Capabilities Index126.91Source
GPT-4o miniOpenAIEpoch Capabilities Index126.58Source
GPT-4 Turbo (Nov 2023)OpenAIEpoch Capabilities Index126.46Source
Llama 3.1-70BMeta AIEpoch Capabilities Index125.94Source
GPT-4 (Mar 2023)OpenAIEpoch Capabilities Index125.88Source
Llama 3.2 90BMeta AIEpoch Capabilities Index125.51Source
Qwen2-72BAlibabaEpoch Capabilities Index125.28Source
DeepSeek-V2 (MoE-236B, May 2024)DeepSeekEpoch Capabilities Index124.78Source
Amazon Nova ProAmazonEpoch Capabilities Index123.8Source
Gemma 3 12BGoogle DeepMindEpoch Capabilities Index123.49Source
GPT-4 (Jun 2023)OpenAIEpoch Capabilities Index123.16Source
Llama 3-70BMeta AIEpoch Capabilities Index122.94Source
Gemini 1.5 Flash (May 2024)Google DeepMindEpoch Capabilities Index122.61Source
Gemma 2 27BGoogle DeepMindEpoch Capabilities Index122.1Source
Mixtral 8x22BMistral AIEpoch Capabilities Index122.05Source
Mistral LargeMistral AIEpoch Capabilities Index122.03Source
phi-3-small 7.4BMicrosoftEpoch Capabilities Index121.85Source
phi-3-medium 14BMicrosoftEpoch Capabilities Index121.25Source
Claude 3 SonnetAnthropicEpoch Capabilities Index120.72Source
Claude InstantAnthropicEpoch Capabilities Index120.31Source
Claude 2AnthropicEpoch Capabilities Index120.15Source
Gemma 2 9BGoogle DeepMindEpoch Capabilities Index119.85Source
Qwen2.5-Coder-32BAlibabaEpoch Capabilities Index119.52Source
Command R+Cohere,Cohere Labs (formerly Cohere for AI)Epoch Capabilities Index119.37Source
Claude 2.1AnthropicEpoch Capabilities Index119.3Source
Mistral NeMoMistral AIEpoch Capabilities Index118.71Source
GPT-3.5 Turbo (Nov 2023)OpenAIEpoch Capabilities Index118.58Source
Qwen2.5-7BAlibabaEpoch Capabilities Index118.53Source
Mixtral 8x7BMistral AIEpoch Capabilities Index118.5Source
Claude 3 HaikuAnthropicEpoch Capabilities Index118.39Source
Ministral 3BMistral AIEpoch Capabilities Index118.12Source
Yi-34B01.AIEpoch Capabilities Index117.43Source
phi-3-mini 3.8BMicrosoftEpoch Capabilities Index117.4Source
Stable Beluga 2Stability AIEpoch Capabilities Index117.16Source
Gemini 1.0 ProGoogle DeepMindEpoch Capabilities Index117.07Source
Llama 3.1-8BMeta AIEpoch Capabilities Index116.61Source
Llama 3-8BMeta AIEpoch Capabilities Index116.49Source
Qwen2.5-Coder-14BNot suppliedEpoch Capabilities Index116.44Source
Gemma 3 4BGoogle DeepMindEpoch Capabilities Index116.01Source
GPT-3.5 Turbo (Jan 2024)OpenAIEpoch Capabilities Index115.74Source
PaLM 2-LNot suppliedEpoch Capabilities Index115.07Source
Llama 2-70BMeta AIEpoch Capabilities Index113.84Source
GPT-3.5 Turbo (Jun 2023)OpenAIEpoch Capabilities Index113.42Source
Qwen2.5-Coder (7B)AlibabaEpoch Capabilities Index113.2Source
Qwen-14BAlibabaEpoch Capabilities Index113.09Source
Mistral 7B v0.1Mistral AIEpoch Capabilities Index112.27Source
Falcon-180BTechnology Innovation InstituteEpoch Capabilities Index112.19Source
internlm-20bNot suppliedEpoch Capabilities Index112.15Source
Gemma 7BGoogle DeepMindEpoch Capabilities Index112.05Source
DeepSeek LLM 67BDeepSeekEpoch Capabilities Index110.47Source
LLaMA-65BMeta AIEpoch Capabilities Index110.23Source
Falcon 2 11BTechnology Innovation InstituteEpoch Capabilities Index109.63Source
Mistral 7B v0.3Mistral AIEpoch Capabilities Index109.06Source
DeepSeek-Coder-V2-Lite-BaseNot suppliedEpoch Capabilities Index108.96Source
PaLM 2-MNot suppliedEpoch Capabilities Index108.37Source
Phi-2MicrosoftEpoch Capabilities Index108.02Source
Qwen2.5-Coder-3BNot suppliedEpoch Capabilities Index107.82Source
Nemotron-4 15BNvidiaEpoch Capabilities Index107.8Source
Yi-9BNot suppliedEpoch Capabilities Index107.69Source
LLaMA-33BMeta AIEpoch Capabilities Index107.49Source
Qwen-7BAlibabaEpoch Capabilities Index106.91Source
PaLM 2-SNot suppliedEpoch Capabilities Index106.26Source
Llama 2-13BMeta AIEpoch Capabilities Index106.25Source
Llama 2-34BMeta AIEpoch Capabilities Index105.29Source
StarCoder 2 15BHugging Face,ServiceNow,Nvidia,BigCodeEpoch Capabilities Index105.09Source
Yi 6B01.AIEpoch Capabilities Index104.86Source
Falcon-40BTechnology Innovation InstituteEpoch Capabilities Index104.54Source
Baichuan2-13BBaichuanEpoch Capabilities Index103.29Source
Qwen2.5-Coder (1.5B)AlibabaEpoch Capabilities Index103.09Source
internlm-7bNot suppliedEpoch Capabilities Index102.97Source
Llama 3.2 1BMeta AIEpoch Capabilities Index102.04Source
INTELLECT-1Prime Intellect,Hugging Face,Arcee AIEpoch Capabilities Index100.97Source
MPT-30BMosaicMLEpoch Capabilities Index100.76Source
LLaMA-13BMeta AIEpoch Capabilities Index100.69Source
Llama 2-7BMeta AIEpoch Capabilities Index99.18Source
chatglm2-6bNot suppliedEpoch Capabilities Index99.05Source
LLaMA-7BMeta AIEpoch Capabilities Index96.82Source
Baichuan 2-7BBaichuanEpoch Capabilities Index96.45Source
DeepSeek Coder 33BDeepSeek,Peking UniversityEpoch Capabilities Index96.45Source
Falcon-7BTechnology Innovation InstituteEpoch Capabilities Index95.23Source
CodeQwen1.5-7BNot suppliedEpoch Capabilities Index94.99Source
vicuna-13b-v1.1Not suppliedEpoch Capabilities Index94.8Source
MPT-7BMosaicMLEpoch Capabilities Index94.73Source
Gemma 2BGoogle DeepMindEpoch Capabilities Index94.34Source
StarCoder 2 7BHugging Face,ServiceNow,Nvidia,BigCodeEpoch Capabilities Index93.7Source
XGen-7BSalesforceEpoch Capabilities Index93.54Source
Qwen-1_8BNot suppliedEpoch Capabilities Index93.07Source
open_llama_7bNot suppliedEpoch Capabilities Index91.8Source
Phi-1.5MicrosoftEpoch Capabilities Index91.68Source
Baichuan1-7BBaichuanEpoch Capabilities Index90.61Source
RedPajama-INCITE-7B-BaseNot suppliedEpoch Capabilities Index90.58Source
Dolly 2.0-12bDatabricksEpoch Capabilities Index89.83Source
DeepSeek Coder 6.7BDeepSeek,Peking UniversityEpoch Capabilities Index89.77Source
StarCoder 2 3BHugging Face,ServiceNow,Nvidia,BigCodeEpoch Capabilities Index88.92Source
Qwen2.5-Coder-0.5BNot suppliedEpoch Capabilities Index88.31Source
Cerebras-GPT-13BCerebras SystemsEpoch Capabilities Index83.42Source
DeepSeek Coder 1.3BDeepSeek,Peking UniversityEpoch Capabilities Index63.88Source
stablelm-tuned-alpha-7bNot suppliedEpoch Capabilities Index56.05Source

A score is meaningful within its benchmark and evaluation settings. The ECI is an index, not a percentage of workplace tasks completed.

Measurement library

Included
benchmark tables.

External benchmark

GPQA diamond

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

FrontierMath-2025-02-28-Private

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

FrontierMath-Tier-4-v2-Private

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

MATH level 5

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

OTIS Mock AIME 2024-2025

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

SWE-Bench verified

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

SimpleQA Verified

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

Chess Puzzles

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

EBR-bench

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

Mystery Game Puzzles

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

MirrorCode

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

Furniture Assembly

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

External benchmark

SpireClimb

Epoch-administered evaluation. See source method and run settings; scores are not interchangeable across tasks.

And when the model
joins the workflow?

A general capability result cannot tell us whether a tender deadline stayed current across four workers, or whether an approval refers to the document about to be sent.

Inspect our operational evaluations