Claude Opus 4.8 (Adaptive Reasoning, Max Effort)
Context 1M tokens
Top 5 cheapest providers (Adjusted $/task)
The same list price can give a different adjusted $/task (caching, token efficiency) — click a price for its inputs.
Within the active global provider, residency and confidentiality filters.
| # | Provider | Platform | Raw input $/1M | Raw output $/1M | Adjusted $/task |
|---|---|---|---|---|---|
| 1 | Amazon Bedrock | OpenRouter | $5.00 | $25.00 | |
| 2 | Azure | OpenRouter | $5.00 | $25.00 | |
| 3 | OpenRouter | $5.00 | $25.00 | ||
| 4 | Google Vertex AI | Google Vertex AI | $5.00 | $25.00 | |
| 5 | AWS Bedrock | AWS Bedrock | $5.50 | $27.50 |
Composite
7 of 7 inputs · 4 from the model family6 radar axes: DesignArena's two boards share one88.2
AA Coding 74.3Coding Agent v1.4 62.1AA Intelligence 42.0AA Agentic —Epoch ECI 158.3Software ECI 160.1DesignArena 1246/1257
Radar: percentile among all models measured on each input; a gap means not measured.
Benchmark sheet
52 of 140 registered benchmark versions · bars show the percentile among all models measured on each benchmark.
- Epoch ECI · attached
- Software ECI · attached
- DesignArena Web Apps (agentic) · attached
- DesignArena Full-Stack · attached
Agentic
AA-AnalystAgent7345.0%
AA-AnalystAgent published 2026-09-10 ↗ · Published board — Tests analyst tasks using agentic Python execution across fourteen domains.
AA-Briefcase761,316 Elo
AA-Briefcase published 2026-09-10 ↗ · Published board — Tests multi-week professional knowledge-work projects with linked tasks and large source collections.
GDPval-AA v2861,489 Elo
GDPval-AA v2 v2 ↗ · Published board — Tests professional knowledge-work deliverables across occupations using AA's Stirrup harness.
Harvey LAB-AA7691.1%
Harvey LAB-AA published 2026-09-10 ↗ · Published board — Tests legal-work deliverables across practice areas on Harvey's private task set. Artificial Analysis' run, graded by one LLM judge against task rubrics — not the same run or scale as Vals AI's HLAB row.
Terminal-Bench Hard (AA)9858.3%
Terminal-Bench Hard (AA) pinned revision 74221fb ↗ · Published board — Tests a pinned 44-task hard subset of terminal-based work using Terminus 2.
Terminal-Bench v2.1 (AA)8984.6%
Terminal-Bench v2.1 (AA) v2.1 ↗ · Published board — Tests terminal-based work on the 89-task verified refresh using Terminus 2.
Terminal-Bench v4.0 (AA)8421.7%
Terminal-Bench v4.0 (AA) v4.0 ↗ · Published board — Tests terminal-based work on the 66-task release using mini-SWE-agent v2.4.6.
ApprenticeBench API (NeoCognition)6228.0%
ApprenticeBench API (NeoCognition) published 2026-09-14 ↗ · Claude Code — An agent completes the same 100 sequential accounts-payable tasks in the Odoo ERP with diminishing mentoring, using dedicated application API calls instead of the screen interface.
Vals Index v2 (Vals AI)7760.9%
Vals Index v2 (Vals AI) v2 ↗ · Published board — GDP-weighted average of agentic model performance across finance, coding and legal tasks.
Finance Agent v2 (Vals Index v2)5353.9%
Finance Agent v2 (Vals Index v2) v2 ↗ · Published board — Multi-step financial reasoning tasks.
Terminal-Bench 2.1 (Vals Index v2)5071.9%
Terminal-Bench 2.1 (Vals Index v2) v2 ↗ · Published board — Command-line interface problem solving, as run by Vals AI.
HLAB — Harvey's Legal Agent Benchmark (Vals Index v2)729.6%
HLAB — Harvey's Legal Agent Benchmark (Vals Index v2) v2 ↗ · Published board — Long-horizon legal work product creation. Vals AI's run of Harvey's Legal Agent Benchmark, reported as accuracy — not comparable with Artificial Analysis' Harvey LAB-AA row.
EBR-bench (Earthborne Rangers, Epoch AI)4628.6%
EBR-bench (Earthborne Rangers, Epoch AI) published 2026-09-18 ↗ · Published board — Learning-capability test: models repeatedly play the obscure campaign board game Earthborne Rangers with note-taking, and the score measures whether results improve across playthroughs.
PostTrainBench v1.13332.9%
PostTrainBench v1.1 ↗ · Claude Code — Agents retrain four small base models (gemma-3-4b-pt, SmolLM3-3B-Base, Qwen3-1.7B-Base, Qwen3-4B-Base) for each of 7 benchmark families; the leaderboard value is the weighted mean over base models × benchmarks, aggregated over the 2–3 listed runs per agent.
Coding
Artificial Analysis Coding Agent Index v1.46462.1%
Artificial Analysis Coding Agent Index v1.4 v1.4 ↗ · Claude Code — The retained AA Coding Agent Index measures coding-agent systems using the earlier three-component implementation.
SciCode (AA subproblems) v1.0.17854.4%
SciCode (AA subproblems) v1.0.1 ↗ · Published board — Tests scientific Python programming with scientist-annotated background information.
SWE-bench Multilingualno percentile84.4%
SWE-bench Multilingual published 2026-09-10 ↗ · Published board — A 300-instance SWE-bench variant with tasks from 42 repositories across 9 programming languages.
SWE-bench Multimodalno percentile38.4%
SWE-bench Multimodal published 2026-09-10 ↗ · Published board — A 480-instance SWE-bench variant whose issue descriptions include visual elements.
SWE-bench Verifiedno percentile88.6%
SWE-bench Verified published 2026-09-10 ↗ · Published board — A 500-instance human-filtered subset of SWE-bench created with OpenAI, served as the default Verified leaderboard view.
Vibe Code Bench (Vals Index v2)8482.7%
Vibe Code Bench (Vals Index v2) v2 ↗ · Published board — End-to-end app-building tasks.
Code Migration (Vals Index v2 subset)7846.9%
Code Migration (Vals Index v2 subset) v2 ↗ · Published board — Porting projects to another language, including COBOL modernization.
FrontierCode 1.1 Main (Cognition) v1.1no percentile46.5%
FrontierCode 1.1 Main (Cognition) v1.1 ↗ · claude-code — Whether a maintainer would merge the agent's pull request, on tasks crafted by open-source maintainers and graded with tests, rubrics and verifiers.
DeepSWE (Datacurve, via Epoch AI)4859.0%
DeepSWE (Datacurve, via Epoch AI) published 2026-09-15 ↗ · mini-swe-agent — Pass rate of coding agents on original, long-horizon software engineering tasks, run by Datacurve with the mini-swe-agent harness.
KernelBench-CUDA: GLM-5.2 Fused MoE (RTX PRO 6000) vrtx-pro-600006.53
KernelBench-CUDA: GLM-5.2 Fused MoE (RTX PRO 6000) vrtx-pro-6000 ↗ · Published board — Fuse GLM-5.2’s MoE forward pass (256+shared experts, top-8 routing) as one CUDA kernel over a frozen shape sweep; Triton, vLLM and other DSLs are banned — CUDA/PTX/CUTLASS only.
KernelBench-CUDA: DeepSeek NSA (RTX PRO 6000) vrtx-pro-60006017.8
KernelBench-CUDA: DeepSeek NSA (RTX PRO 6000) vrtx-pro-6000 ↗ · Published board — Implement DeepSeek’s Native Sparse Attention block (block top-n routing + sparse attention) as a CUDA kernel over a frozen shape sweep.
KernelBench-CUDA: MegaQwen Decode (RTX PRO 6000) vrtx-pro-600000.97
KernelBench-CUDA: MegaQwen Decode (RTX PRO 6000) vrtx-pro-6000 ↗ · Published board — Improve the known MegaQwen CUDA megakernel geometry for decode-only token throughput at context lengths 2k–128k; prefill is untimed.
KernelBench-CUDA: Grid MinGRU SPS (RTX PRO 6000) vrtx-pro-6000032.7
KernelBench-CUDA: Grid MinGRU SPS (RTX PRO 6000) vrtx-pro-6000 ↗ · Published board — Non-LLM RL simulation: maximise simulation steps per second for a grid world with a MinGRU agent (roofline anchored at 150M peak SPS); fusion optional.
AA Coding Index8874.3
AA Coding Index published 2026-09-19 ↗ · Published board — Artificial Analysis publishes this board without a version number; we keep the date each result was retained.
Efficiency
ApprenticeBench API cost per task (NeoCognition)23$4.59
ApprenticeBench API cost per task (NeoCognition) published 2026-09-14 ↗ · Claude Code — ApprenticeBench's published USD cost per task for each API-board model-harness-effort configuration on the 100-task accounts-payable job.
Vals Index v2 cost per test (Vals AI)20$12.67
Vals Index v2 cost per test (Vals AI) v2 ↗ · Published board — Vals AI's published USD cost per test for each model on the Vals Index v2.
FrontierCode 1.1 Main cost per rollout (Cognition) v1.1no percentile$9.62
FrontierCode 1.1 Main cost per rollout (Cognition) v1.1 ↗ · claude-code — Cognition's published mean USD spend per rollout for each FrontierCode 1.1 Main model and reasoning effort.
Instruction-following
IFBench (AA single-turn)7362.2%
IFBench (AA single-turn) published 2026-09-10 ↗ · Published board — Tests precise single-turn instructions with deterministic rule checks.
Knowledge
Humanity's Last Exam (AA text-only)9748.7%
Humanity's Last Exam (AA text-only) published 2026-09-10 ↗ · Published board — Tests expert-level knowledge on AA's text-only Humanity's Last Exam subset.
AA-Omniscience Index9628.8
AA-Omniscience Index published 2026-09-10 ↗ · Published board — Tests factual reliability while rewarding correct answers and penalizing hallucinations.
Legal Research Bench (Vals Index v2)7743.8%
Legal Research Bench (Vals Index v2) v2 ↗ · Published board — Case and statute research with citation-backed answers.
SimpleQA Verified (run by Epoch AI)6453.0%
SimpleQA Verified (run by Epoch AI) published 2026-09-16 ↗ · Published board — Short fact-seeking questions answered without tools, testing whether a model knows a fact rather than guessing; Google DeepMind's cleaned version of OpenAI's SimpleQA, run by Epoch AI.
Long-context
GDP.pdf (AA)8522.8%
GDP.pdf (AA) published 2026-09-10 ↗ · Published board — Tests professional reasoning over long PDFs with AA document preparation and grading.
AA-LCR v1.18277.7%
AA-LCR v1.1 v1.1 ↗ · Published board — Tests reasoning across multiple long documents with corrected answer keys and grading.
MLCR-AA8845.6%
MLCR-AA published 2026-09-10 ↗ · Published board — Tests medical-record synthesis and reasoning across long, fragmented case documents.
Math
ArXivMath 06/2026 (MathArena) v2026-063070.0%
ArXivMath 06/2026 (MathArena) v2026-06 ↗ · Published board — Research-level math problems with a checkable final answer, drawn from arXiv papers submitted in June 2026, so they postdate most training data.
BrokenArXiv 06/2026 (MathArena) v2026-063041.7%
BrokenArXiv 06/2026 (MathArena) v2026-06 ↗ · Published board — Plausible but false proof statements taken from June 2026 arXiv papers; a model scores by refusing to prove them and saying the statement is false as written.
FrontierMath Tiers 1–3 v2 (Epoch AI)7380.0%
FrontierMath Tiers 1–3 v2 (Epoch AI) v2 ↗ · Published board — Unpublished, expert-written mathematics problems from undergraduate to research level with automatically checkable answers, run by Epoch AI on its private v2 set.
FrontierMath Tier 4 v2 (Epoch AI)5956.1%
FrontierMath Tier 4 v2 (Epoch AI) v2 ↗ · Published board — The hardest, research-level tier of Epoch AI's unpublished FrontierMath problems, run by Epoch AI on its private v2 set.
Reasoning
Chess Puzzles (Epoch AI)6634.0%
Chess Puzzles (Epoch AI) published 2026-09-18 ↗ · Published board — Best-move selection on 100 novel chess positions generated programmatically by Epoch AI, each with a single Stockfish-verified best move; probes spatial reasoning and planning.
Mystery Game Puzzles (Epoch AI)7336.0%
Mystery Game Puzzles (Epoch AI) published 2026-09-18 ↗ · Published board — Best-move selection on 100 mid-game positions of a well-known game whose identity Epoch deliberately keeps undisclosed, generated programmatically like Chess Puzzles.
AA Intelligence Index9642.0
AA Intelligence Index published 2026-09-19 ↗ · Published board — Artificial Analysis publishes this board without a version number; we keep the date each result was retained.
Science
CritPt (AA)9420.9%
CritPt (AA) published 2026-09-10 ↗ · Published board — Tests research-level physics reasoning with Python, symbolic and numerical answers.
GPQA Diamond (AA)9392.0%
GPQA Diamond (AA) published 2026-09-10 ↗ · Published board — Tests graduate-level biology, physics and chemistry knowledge on the Diamond subset.
GPQA Diamond (Epoch AI run)7491.0%
GPQA Diamond (Epoch AI run) published 2026-09-18 ↗ · Published board — Epoch AI's own inspect-ai runs of GPQA Diamond, the 198-question graduate-level science multiple-choice set.
Tool-use
AutomationBench-AA v1.0.66945.6%
AutomationBench-AA v1.0.6 ↗ · Published board — Tests multi-app SaaS workflows through REST tools on a held-out AutomationBench split.
EnterpriseOps-Gym-AA6844.0%
EnterpriseOps-Gym-AA published 2026-09-10 ↗ · Published board — Tests enterprise workflows through MCP tools against resettable application environments.
τ²-Bench Telecom (AA)9394.4%
τ²-Bench Telecom (AA) published 2026-09-10 ↗ · Published board — Tests dual-control telecom agents that coordinate tool use with a simulated user.
τ³-Banking (AA) v1.0.17434.2%
τ³-Banking (AA) v1.0.1 ↗ · Published board — Tests banking support agents that retrieve policies and change account state through tools.
Excel Modeling Benchmark (Vals Index v2)7869.4%
Excel Modeling Benchmark (Vals Index v2) v2 ↗ · Published board — Building and editing financial models in spreadsheets.
Unusual results
1 threshold-crossing signal flagged · 38 eligible benchmark families
How flags are calculated
Heuristic screen, not statistical significance: benchmark families are correlated and source uncertainty is unknown. Peer evidence needs ≥ 20 independently measured matched configurations from ≥ 10 distinct model families. A flag needs a directed population z-score of magnitude ≥ 1.5 and a gap of ≥ 1.5 from the leave-one-benchmark-family-out mean z in the same direction, over ≥ 5 other benchmark families.
- unusually strong Humanity's Last Exam (AA text-only) published 2026-09-10
Why
Observed: 0.48656 fractionPeer mean 0.15283 · peer sd 0.14006 · n 608 · families 449Directed z 2.383 · baseline z 0.86 · gap 1.522 · profile n 370.48656 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗Evidence
Axis: Humanity's Last Exam (AA text-only) · published 2026-09-10 · Published boardExact value:0.486561631139944fractionObserved: 2026-09-10T21:47:16.627Z · publication date: not recordedObservation id:aa:992b7b84-5069-4c6a-9295-834252553d50:hle
Missing coverage · 91 benchmark versions
No result does not mean a zero, or that the model was never tested. Collection failures and disputed versions retain their distinct status.
Variants / reasoning settings
Artificial Analysis snapshot 2026-09-19 · Data: Artificial Analysis · Data: Epoch AI (CC BY)
SUBSCRIPTION PLAN
GitHub Copilot
Current usage-based billing · model token cost is converted to AI Credits at 1 credit = $0.01.
Fast mode research preview · separate opt-in mode, not the standard model price.
Legacy annual Pro/Pro+ request billing only.
Legacy annual-plan multiplier.
Token offers by platform · 12 offers (Adjusted $/task)
Click any underlined price to see how it is estimated and where each input comes from. How we calculate adjusted cost.
12 offers within the active global filters; “—” means the catalog is active but no public token price is available.
OpenRouter (8)
| Amazon Bedrock | globalamazon-bedrock | $5.00 raw in $/1M | $25.00 raw out $/1M | |
| Azure | globalazure/global | $5.00 raw in $/1M | $25.00 raw out $/1M | |
| globalgoogle-vertex/global | $5.00 raw in $/1M | $25.00 raw out $/1M | ||
| Amazon Bedrock | globalamazon-bedrock/us | $5.50 raw in $/1M | $27.50 raw out $/1M | |
| globalgoogle-vertex/us | $5.50 raw in $/1M | $27.50 raw out $/1M | ||
| eugoogle-vertex/europeEU | $5.50 raw in $/1M | $27.50 raw out $/1M | ||
| Azure | globalazure/us | $5.00 raw in $/1M | $25.00 raw out $/1M | |
| Amazon Bedrock | euamazon-bedrock/eu-west-1EU | $5.50 raw in $/1M | $27.50 raw out $/1M |
Google Vertex AI (2)
| Google Vertex AI | global | $5.00 raw in $/1M | $25.00 raw out $/1M | |
| Google Vertex AI | euEU | $5.50 raw in $/1M | $27.50 raw out $/1M |
AWS Bedrock (1)
| AWS Bedrock | eu-central-1 (EU cross-region inference profile)EU | $5.50 raw in $/1M | $27.50 raw out $/1M |
T-Systems LLM Hub (1)
| T-Systems LLM Hub | euEU | $5.67 raw in $/1M | $28.36 raw out $/1M |