Pass IndexThe State of AISign in

Claude Opus 4

Claude Opus 4 is Anthropic's most powerful model yet and the state-of-the-art coding model.

text + image + file → text · made by Anthropic

$15$75per Mtok in / out
Anthropic · 12 sellers

Sold by 14 ways

SellerLaneRate
Snowflake Cortex cloudstandard$5per Mtok in$25per Mtok out$0.5per Mtok cachedraw.githubusercontent.com · read 2026-09-16
Anthropic apibatch$7.5per Mtok in$37.5per Mtok outplatform.claude.com · read 2026-08-24
AWS Bedrock cloudstandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writeraw.githubusercontent.com · read 2026-09-16
Anthropic apistandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writeplatform.claude.com · read 2026-08-24
DigitalOcean cloudstandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writemodels.dev · read 2026-09-16
Google Vertex AI cloudstandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writemodels.dev · read 2026-09-16
Databricks cloudstandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writeraw.githubusercontent.com · read 2026-09-16
Requesty aggregatornot on the seller's list todayvertex us-east5$13.5per Mtok in$67.5per Mtok out$1.35per Mtok cachedrouter.requesty.ai · read 2026-08-25
Requesty aggregatorvertex$15per Mtok in$75per Mtok out$1.5per Mtok cachedrouter.requesty.ai · read 2026-08-28
ElectronHub aggregatorstandard$15per Mtok in$75per Mtok outapi.electronhub.ai · read 2026-09-16
Nous Research aggregatorstandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writeinference-api.nousresearch.com · read 2026-09-12
OpenRouter aggregatorstandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writeopenrouter.ai · read 2026-08-24
Vercel AI Gateway aggregatorstandard$15per Mtok in$75per Mtok out$1.5per Mtok cached$18.75per Mtok cache writeai-gateway.vercel.sh · read 2026-08-25
DeepInfra aggregatorstandard$16.5per Mtok in$82.5per Mtok outapi.deepinfra.com · read 2026-08-25

Measured 47 standings

PlaceBoardMetricScore
8thof 22Cybench — Epoch AIUnguided % Solved0.38
8thof 49Lech mazur writing — Epoch AIMean score8.36
12thof 262MMLU-ProOverall (accuracy) (thinking)0.873
14thof 69Aider polyglot coding leaderboardPercent correct72
14thof 81Forecastbench — Epoch AIOverall score61.1
17thof 72Aider polyglot — Epoch AIPercent correct72
17thof 18Algotune — Epoch AIScore1.33
17thof 47SWE-bench Verified (default "Bash Only" view, agent = mini-SWE-agent)% Resolved67.6
18thof 50Metr time horizons — Epoch AIaverage_score0.639
19thof 28LiveCodeBench Leaderboard (default window 8/1/2024–5/1/2025, 454 problems)Pass@156.6
19thof 28LiveCodeBench Leaderboard (default window 8/1/2024–5/1/2025, 454 problems)Pass@1 (thinking)56.6
22ndof 38Gso — Epoch AIScore OPT@10.069
23rdof 180SWE-bench Verified (Bash Only toggle off — all agents)% Resolved73.2
24thof 38Vpct — Epoch AICorrect0.38
25thof 46Enigma eval — Epoch AIAccuracy5.57
26thof 41Deepresearchbench — Epoch AIAverage score0.468
26thof 35Swe bench verified — Epoch AImean_score70.661
29thof 51Humanity's Last Exam (Epoch AI replication)Accuracy10.72
31stof 108Math level 5 — Epoch AImean_score85.045
32ndof 42Fictionlivebench — Epoch AI120k token score0.375
32ndof 34LMArena · SearchScore (Elo)1126
34thof 102Simplebench — Epoch AIScore (AVG@5)0.588
46thof 72Frontiermath tier 4 — Epoch AImean_score4.167
46thof 123Lmca — Epoch AIScore37.43
60thof 161Dtbench — Epoch AIAccuracy81.6
70thof 101FrontierMath (Tiers 1-3)mean_score4.483
71stof 109Design Arena · svg (models)Elo1152
74thof 152LMArena · VisionScore (Elo)1207
76thof 105Vectara Hallucination LeaderboardHallucination Rate12
78thof 155Design Arena · gamedev (models)Elo1193
81stof 147Design Arena · 3d (models)Elo1167
91stof 150Design Arena · uicomponent (models)Elo1168
92ndof 170Weirdml — Epoch AIAccuracy43.72
96thof 156Design Arena · codecategories (models)Elo1176
98thof 162Design Arena · website (models)Elo1177
99thof 152Design Arena · dataviz (models)Elo1161
107thof 402LMArena · TextScore (Elo)1426
121stof 232ARC-AGI-2 (Semi-Private)Score (%)8.61
123rdof 177Critpt — Epoch AIAccuracy0.3
124thof 227Arc agi 2 — Epoch AIScore8.61
130thof 205ARC-AGI-1 (Public Eval)Score (%)54.25
140thof 216ARC-AGI-2 (Public Eval)Score (%)3.33
144thof 313GPQA Diamondmean_score76.263
150thof 290OTIS Mock AIME 2024-2025mean_score64.444
158thof 230ARC-AGI-1 (Semi-Private)Score (%)35.67
168thof 244Arc agi — Epoch AIScore35.67
367thof 656Epoch capabilities index — Epoch AIECI Score142.7

About

Claude Opus 4 — a text model from Anthropic, sold by 12 companies from $5 in and $25 out per million tokens, placed 12th of 262 on MMLU-Pro.

It takes text, images and files and returns text, with a context window of 200,000 tokens. It was published in May 2025, trained on material up to March 2025. Its sellers say it can reason step by step and call a tool. The catalogue files it under reasoning, code and agents. Twelve companies sell it. The cheapest is $5 in and $25 out per million tokens at Snowflake Cortex. Beside the standard rate there are batch and separately routed lanes. It has been measured on 46 boards, and stands best at 12th of 262 on MMLU-Pro.

Every current figure

Maker
Anthropic
Register
model
Takes
text + image + file
Returns
text
Context
200,000 tokens
Longest answer
32,000 tokens
Published
May 2025
Knowledge to
March 2025
Licence
not read
Sellers
12
Price
$5 in and $25 out per million tokens — Snowflake Cortex
Boards
46
Best place
12th of 262 — MMLU-Pro

Known as 11 names

claude-opus-4-20250514 (32k thinking)claude-opus-4-20250514 (no think)Claude-Opus-4Claude-Opus-4 (Thinking)claude-opus-4-searchClaude-4-Opus-ThinkingClaude 4 Opusanthropic/claude-4-opusanthropic/claude-opus-4vertex/claude-opus-4vertex/claude-opus-4@us-east5