Leaderboard 638 current results

Select models with +, then open Compare.

Leaderboard
#	Model	Score	Config	Evaluated	Source	Actions
601	#601jamba-1-5-miniAI21 Labs	5.21	aa_slug=jamba-1-5-mini · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
602	#602lfm2-2-6bLiquid AI	5.21	aa_slug=lfm2-2-6b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
603	#603qwen3-1.7b-instruct-reasoningAlibaba Group	5.19	aa_slug=qwen3-1.7b-instruct-reasoning · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
604	#604granite-4-0-h-nano-1bIBM	5.19	aa_slug=granite-4-0-h-nano-1b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
605	#605Qwen3 8BQwen	5.16	aa_slug=qwen3-8b-instruct-reasoning · version=4.3 · estimated=false	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
606	#606jamba-1-6-miniAI21 Labs	5.16	aa_slug=jamba-1-6-mini · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
607	#607Mixtral 8x7BMistral AI	5.12	aa_slug=mixtral-8x7b-instruct · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
608	#608gemma-3-270mGoogle	5.12	aa_slug=gemma-3-270m · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
609	#609apertus-70b-instructSwiss AI Initiative	5.11	aa_slug=apertus-70b-instruct · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
610	#610Granite 4.0 MicroIBM	5.11	aa_slug=granite-4-0-micro · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
611	#611deephermes-3-llama-3-1-8b-previewNous Research	5.08	aa_slug=deephermes-3-llama-3-1-8b-preview · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
612	#612Mistral 7BMistral AI	5.03	aa_slug=mistral-7b-instruct · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
613	#613command-r-03-2024Cohere	5.03	aa_slug=command-r-03-2024 · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
614	#614llama-65bMeta AI	5.03	aa_slug=llama-65b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
615	#615qwen-chat-14bAlibaba Group	5.03	aa_slug=qwen-chat-14b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
616	#616claude-instantAnthropic	5.03	aa_slug=claude-instant · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
617	#617granite-4-0-nano-1bIBM	5.01	aa_slug=granite-4-0-nano-1b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
618	#618Molmo2-8BAllen Institute for AI	5	aa_slug=molmo2-8b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
619	#619lfm2-8b-a1bLiquid AI	4.93	aa_slug=lfm2-8b-a1b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
620	#620granite-3-3-8b-instructIBM	4.92	aa_slug=granite-3-3-8b-instruct · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
621	#621qwen3-1.7b-instructAlibaba Group	4.86	aa_slug=qwen3-1.7b-instruct · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
622	#622Gemma 3 27BGoogle	4.85	aa_slug=gemma-3-27b · version=4.3 · estimated=false	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
623	#623Ministral 3 3BMistral AI	4.84	aa_slug=ministral-3-3b · version=4.3 · estimated=false	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
624	#624gemma-3n-e4bGoogle	4.83	aa_slug=gemma-3n-e4b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
625	#625granite-4-0-h-350mIBM	4.83	aa_slug=granite-4-0-h-350m · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
626	#626Gemma 3 4BGoogle	4.83	aa_slug=gemma-3-4b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
627	#627granite-4-0-350mIBM	4.83	aa_slug=granite-4-0-350m · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
628	#628llama-3-instruct-8bMeta AI	4.83	aa_slug=llama-3-instruct-8b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
629	#629gemma-3-1bGoogle	4.83	aa_slug=gemma-3-1b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
630	#630tiny-aya-globalCohere	4.83	aa_slug=tiny-aya-global · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
631	#631qwen3-0.6b-instructAlibaba Group	4.83	aa_slug=qwen3-0.6b-instruct · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
632	#632Llama-3.2-1BMeta AI	4.83	aa_slug=llama-3-2-instruct-1b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
633	#633LFM2-1.2BLiquid AI	4.83	aa_slug=lfm2-1-2b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
634	#634LFM2.5-VL-1.6BLiquid AI	4.83	aa_slug=lfm2-5-vl-1-6b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
635	#635Qwen3-0.6BQwen	4.83	aa_slug=qwen3-0.6b-instruct-reasoning · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
636	#636apertus-8b-instructSwiss AI Initiative	4.83	aa_slug=apertus-8b-instruct · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
637	#637gemma-3n-e2bGoogle	4.83	aa_slug=gemma-3n-e2b · version=4.3 · estimated=true	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History
638	#638Gemma 3 12BGoogle	3.75	aa_slug=gemma-3-12b · version=4.3 · estimated=false	— (obs. 11 Sept 2026)	artificialanalysis.aiT2	History

Scores are reported as published, with their evaluation configuration (harness, prompting, judge). The bar is relative to the best score on this page. Results with different configs are not directly comparable — see methodology.

The config filter matches a value inside each result's configuration (server-side, `config=` on the API). Chips are the values shared by several rows on the first page; per-model identifiers are not offered.

Definition

Category: composite
Source:AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2observed 8 h agomedium
Task: composite of several evaluations run by Artificial Analysis
Source:AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2observed 8 h agomedium
Metric: index
Source:AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2observed 8 h agomedium
Direction: Higher is better
Known limitations: Proprietary composite; methodology versions change.
Source:AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2observed 8 h agomedium
Website: artificialanalysis.ai
Source:AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2observed 8 h agomedium

Each value shows its source, tier and observation time. Missing rows mean no source stated them. How results are recorded →

Evaluated models: llama-2-chat-7b, grok-3-mini-reasoning, claude-opus-5-medium, gemma-4-12B, deepseek-v4-flash, grok-4-3-low, llama-65b, gemini-2.0-flash-exp +630(638 total)

As of

Rewind the record: see this entity's attributes exactly as AI Atlas knew them on a given day.

Claim history

5 claims · 5 properties

Categorycategory1

Claim history for Category
Value	Valid from → to	Status	Source	Confidence	Extractor
composite	11 Sept 2026 20:50 UTC → current	current	AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2	medium	curated

Known limitationsknown_limitations1

Claim history for Known limitations
Value	Valid from → to	Status	Source	Confidence	Extractor
Proprietary composite; methodology versions change.	11 Sept 2026 20:50 UTC → current	current	AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2	medium	curated

Metricmetric1

Claim history for Metric
Value	Valid from → to	Status	Source	Confidence	Extractor
index	11 Sept 2026 20:50 UTC → current	current	AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2	medium	curated

Tasktask1

Claim history for Task
Value	Valid from → to	Status	Source	Confidence	Extractor
composite of several evaluations run by Artificial Analysis	11 Sept 2026 20:50 UTC → current	current	AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2	medium	curated

Websitewebsite1

Claim history for Website
Value	Valid from → to	Status	Source	Confidence	Extractor
https://artificialanalysis.ai	11 Sept 2026 20:50 UTC → current	current	AI Atlas curated registry (YAML, versioned in git, every entry carries its source URL)T2	medium	curated

Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →

11 Sept 2026
New benchmarkBenchmarkArtificial Analysis Intelligence Index
New benchmark: Artificial Analysis Intelligence Index
registry
artificialanalysis.ai

Full timeline →

Source documents
Source	Document	Type	Tier	Last observed	Snapshots
Artificial Analysis	artificialanalysis.ai/leaderboards/models	leaderboard	T2· Quality secondary	6 h ago	1

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.