Grok 4.6
Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work.
text + image → text · made by xAI
$2→$6per Mtok in / out
xAI · 15 sellers
Sold by 24 ways
| Seller | Lane | Rate |
|---|---|---|
| AWS Bedrock cloud | flex | $1per Mtok in$3per Mtok outraw.githubusercontent.com · read 2026-10-31 |
| Google Vertex AI cloud | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedmodels.dev · read 2026-10-31 |
| Microsoft Azure AI cloud | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedmodels.dev · read 2026-10-31 |
| Oracle Cloud Infrastructure cloud | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedmodels.dev · read 2026-10-31 |
| xAI api | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cacheddocs.x.ai · read 2026-10-31 |
| AWS Bedrock cloud | standard | $2.2per Mtok in$6.6per Mtok out$0.55per Mtok cachedmodels.dev · read 2026-10-31 |
| Databricks cloud | standard | $2.5per Mtok in$7.5per Mtok out$0.63per Mtok cached$2.5per Mtok cache writeraw.githubusercontent.com · read 2026-10-31 |
| AWS Bedrock cloud | priority | $3.5per Mtok in$10.5per Mtok out$0.88per Mtok cachedraw.githubusercontent.com · read 2026-10-31 |
| Microsoft Azure AI cloud | long-context | $4per Mtok in$12per Mtok out$1per Mtok cachedraw.githubusercontent.com · read 2026-10-31 |
| xAI 0 ways | $4per Mtok in$12per Mtok out$1per Mtok cached | |
| xAI apinot on the seller's list today | long-context | $4per Mtok in$12per Mtok out$1per Mtok cacheddocs.x.ai · read 2026-08-25 |
| AtlasCloud aggregator | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedapi.atlascloud.ai · read 2026-09-20 |
| ElectronHub aggregator | standard | $2per Mtok in$6per Mtok outapi.electronhub.ai · read 2026-10-31 |
| GitHub Copilot aggregator | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedmodels.dev · read 2026-10-31 |
| Inference.net aggregator | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedapi.inference.net · read 2026-10-31 |
| Nous Research aggregator | standard | $2per Mtok in$6per Mtok out$0.000002per image$0.5per Mtok cachedinference-api.nousresearch.com · read 2026-09-12 |
| OpenRouter 5 ways | $2per Mtok in$6per Mtok out$0.5per Mtok cached | |
| OpenRouter aggregator | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedopenrouter.ai · read 2026-08-24 |
| OpenRouter aggregator | amazon-bedrock/us-west-2 | $2.2per Mtok in$6.6per Mtok out$0.55per Mtok cachedopenrouter.ai · read 2026-08-26 |
| OpenRouter aggregator | xai/zdr/us | $2.2per Mtok in$6.6per Mtok out$0.55per Mtok cachedopenrouter.ai · read 2026-09-29 |
| OpenRouter aggregator | xai/priority | $4per Mtok in$12per Mtok out$1per Mtok cachedopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | xai/zdr/priority | $4per Mtok in$12per Mtok out$1per Mtok cachedopenrouter.ai · read 2026-08-25 |
| Requesty aggregator | xai | $2per Mtok in$6per Mtok out$0.5per Mtok cachedrouter.requesty.ai · read 2026-08-28 |
| Vercel AI Gateway aggregator | standard | $2per Mtok in$6per Mtok out$0.5per Mtok cachedvercel.com · read 2026-08-25 |
| Vercel AI Gateway aggregatornot on the seller's list today | sandbox | $2per Mtok in$6per Mtok outvercel.com · read 2026-08-25 |
| Venice AI aggregator | standard | $2.27per Mtok in$6.8per Mtok out$0.57per Mtok cachedmodels.dev · read 2026-10-31 |
Measured 59 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 1stof 54 | Design Arena · androidnative (agents) | Elo | 1290 |
| 7thof 50 | Design Arena · godotgamedev (agents) | Elo | 1276 |
| 8thof 102 | Simplebench — Epoch AI | Score (AVG@5) | 0.759 |
| 9thof 131 | Artificial Analysis · Agentic Index | Index | 53 |
| 9thof 163 | Artificial Analysis · Coding Index | Index | 76.8 |
| 9thof 235 | Dtbench — Epoch AI | Accuracy (xhigh) | 97.33 |
| 10thof 60 | Apex agents — Epoch AI | Pass@1 score | 0.653 |
| 10thof 61 | Design Arena · fullstack (agents) | Elo | 1264 |
| 10thof 36 | Design Arena · htmlslides (agents) | Elo | 1146 |
| 10thof 46 | Frontiercode — Epoch AI | Main score | 0.48 |
| 12thof 92 | Forecastbench — Epoch AI | Overall score (xhigh) | 65.5 |
| 13thof 34 | Design Arena · pdf (agents) | Elo | 1258 |
| 13thof 326 | GPQA Diamond | mean_score (high) | 94.003 |
| 13thof 71 | Vending bench 2 — Epoch AI | Score | 9047.033 |
| 14thof 35 | Blueprint bench 2 — Epoch AI | Score | 33.189 |
| 14thof 28 | Ebr bench — Epoch AI | mean_score (xhigh) | 30.476 |
| 14thof 21 | Frontierswe — Epoch AI | Score (xhigh) | 25.286 |
| 16thof 39 | Design Arena · agenticgamedev (agents) | Elo | 1158 |
| 16thof 39 | Design Arena · blender3d (agents) | Elo | 1156 |
| 17thof 93 | Design Arena · asciiart (models) | Elo | 1273 |
| 17thof 68 | Design Arena · mobileapps (agents) | Elo | 1232 |
| 17thof 41 | Design Arena · python-pptxslides (agents) | Elo | 1199 |
| 18thof 68 | Design Arena · webapps (agents) | Elo | 1236 |
| 18thof 35 | Furniture assembly — Epoch AI | mean_score (xhigh) | 40 |
| 18thof 30 | Mirrorcode — Epoch AI | mean_score (xhigh) | 41.111 |
| 19thof 135 | Webdev arena — Epoch AI | Arena Score (high) | 1617.64 |
| 20thof 69 | Deepswe — Epoch AI | Pass@1 (medium) | 0.675 |
| 20thof 306 | OTIS Mock AIME 2024-2025 | mean_score (xhigh) | 99.167 |
| 25thof 141 | LMArena · WebDev | Score (Elo) (high) | 1617 |
| 26thof 142 | Mystery game puzzles — Epoch AI | mean_score (xhigh) | 34 |
| 28thof 132 | Ale bench — Epoch AI | Performance (xhigh) | 1508.08 |
| 29thof 163 | Design Arena · codecategories (models) | Elo | 1290 |
| 30thof 154 | Design Arena · dataviz (models) | Elo | 1283 |
| 30thof 114 | Design Arena · svg (models) | Elo | 1236 |
| 30thof 93 | Simpleqa verified — Epoch AI | mean_score (high) | 49.3 |
| 31stof 162 | Design Arena · gamedev (models) | Elo | 1295 |
| 31stof 84 | Proofbench — Epoch AI | Accuracy | 51 |
| 34thof 56 | Gdp pdf — Epoch AI | GDP.pdf score (xhigh) | 0.172 |
| 35thof 150 | Design Arena · 3d (models) | Elo | 1273 |
| 35thof 157 | Design Arena · uicomponent (models) | Elo | 1290 |
| 35thof 169 | Design Arena · website (models) | Elo | 1287 |
| 35thof 122 | Frontiermath tiers 1 3 v2 — Epoch AI | mean_score (xhigh) | 65.965 |
| 38thof 173 | Weirdml — Epoch AI | Accuracy (high) | 67.29 |
| 39thof 78 | Frontiermath tier 4 v2 — Epoch AI | mean_score (xhigh) | 31.707 |
| 40thof 82 | Cursorbench — Epoch AI | Score (xhigh) | 41.4 |
| 40thof 235 | Scicode — Epoch AI | Score (high) | 56.481 |
| 45thof 165 | LMArena · Vision | Score (Elo) (high) | 1262 |
| 47thof 53 | LMArena · Agent | Net Improvement (%) | 1.05 |
| 54thof 112 | ARC-AGI-3 (Semi-Private) | Score (%) (xhigh) | 2.11 |
| 55thof 273 | ARC-AGI-1 (Public Eval) | Score (%) (xhigh) | 96.25 |
| 63rdof 154 | Chess puzzles — Epoch AI | mean_score (xhigh) | 33 |
| 66thof 260 | Critpt — Epoch AI | Accuracy (xhigh) | 19.714 |
| 68thof 282 | ARC-AGI-2 (Public Eval) | Score (%) (xhigh) | 68.33 |
| 72ndof 298 | ARC-AGI-2 (Semi-Private) | Score (%) (xhigh) | 67.08 |
| 72ndof 414 | LMArena · Text | Score (Elo) (high) | 1454 |
| 76thof 197 | Lmca — Epoch AI | Score (xhigh) | 48.45 |
| 77thof 290 | Arc agi 2 — Epoch AI | Score (xhigh) | 67.083 |
| 97thof 298 | ARC-AGI-1 (Semi-Private) | Score (%) (medium) | 87.5 |
| 107thof 309 | Arc agi — Epoch AI | Score (medium) | 87.5 |
About
Grok 4.6 — a text model from xAI, sold by 15 companies from $2 in and $6 out per million tokens, placed 1st of 54 on Design Arena · androidnative (agents).
It takes text and images and returns text, with a context window of 500,000 tokens. It was published in August 2026, trained on material up to February 2026. Its sellers say it can reason step by step and call a tool. The catalogue files it under reasoning, code and agents. Fifteen companies sell it. The cheapest is $2 in and $6 out per million tokens at AtlasCloud. Beside the standard rate there are flex, faster, separately routed and regional lanes. It has been measured on 59 boards, and stands best at 1st of 54 on Design Arena · androidnative (agents).
Every current figure
- Maker
- xAI
- Register
- model
- Takes
- text + image
- Returns
- text
- Context
- 500,000 tokens
- Published
- August 2026
- Knowledge to
- February 2026
- Licence
- not read
- Sellers
- 15
- Price
- $2 in and $6 out per million tokens — AtlasCloud
- Boards
- 59
- Best place
- 1st of 54 — Design Arena · androidnative (agents)
Known as 11 names
Grok 4.6 (high)Grok 4.6 (low)Grok 4.6 (medium)Grok 4.6 (xhigh)grok-4.6_xhighgrok-4.6_highgrok-4.6-highxai/grok-4.6x-ai/grok-4.6spacexai/grok-4.6grok-4.6