Grok 4.6 (batch)
Sold by 4 sellers, timed at 62.6 tok/s, scored 31.0 on the Intelligence Index, on 20 Design Arena boards, and with results on 2 openly licensed boards.
retrieved sha256 · 0bb53a77713f
What sellers charge
4 sellers · 2x from cheapest to dearestOne tick per seller, on a log scale of input price per million from $2.00 at xAI to $4.00 at xAI. The $1.60 listed is the cheapest on offer; 4 sellers are dearer. A faint tick declares no number format. Every seller is named on the endpoint ledger below.
Measured speed
2,722 requests · 4 endpoints · 30-minute windowOpenRouter's own timing of its routed traffic, weighted by requests, over one 30-minute window read . Every route is on the endpoint ledger below.
Every endpoint
4 priced · 4 timed · cheapest first- xAIxai$2.00$6.00$0.50not declared450knot matchednonenonenot matched98.95%
- xAIxai/zdr$2.00$6.00$0.50not declared450knot matchednonenonenot matched99.76%
- xAIxai/priority$4.00$12.00$1.00not declared450knot matchednonenonenot matchedno reading
- xAIxai/zdr/priority$4.00$12.00$1.00not declared450knot matchednonenonenot matchedno reading
- xAInot matchednot matchednot matchednot matchedno offer56 tok/s402.9s1,662 reqno offer
- xAInot matchednot matchednot matchednot matchedno offer77 tok/s49.4759ms2 reqno offer
- xAInot matchednot matchednot matchednot matchedno offer73 tok/s391.93s1,057 reqno offer
- xAInot matchednot matchednot matchednot matchedno offer75 tok/s40896ms1 reqno offer
Each track runs from a route's p90, the rate nine in ten of its requests met or beat, up to its median. Weighted by requests, the figure above reports where the traffic went rather than what one route does: 1 of the 4 timed routes are slower than it, and the middle route runs at 74 tok/s. The widest median-to-p90 gap is xAI's, 1.88x.
xAI publishes 4 offers and had 4 routes timed. The two documents share no route id, so those offers and routes are printed apart rather than paired.
Uptime is OpenRouter's own figure for each seller over the trailing day, read once a day, and the column prints the latest reading, taken . 6 of the 6 sellers with a reading fell under 99% on at least one of the 23 days held, 23 Oct 2026 to 21 Nov 2026: xAI (xai/priority) at 0.00% on 7 Nov 2026, xAI (xai/zdr/priority) at 0.00% on 7 Nov 2026, xAI (xai/zdr/us) at 0.00% on 10 Nov 2026, xAI (xai/zdr) at 69.40% on 8 Nov 2026, and 2 more. Source: OpenRouter's public endpoints API, one reading a day for each seller: 1,211 sellers carried one on the latest of the 23 days held.
Source: OpenRouter's public endpoints API. 1257 sellers across 359 models when the mirror ran, 186 of them served by more than one. List prices, not what anybody paid; a discount is the seller's own. Read . Speed is OpenRouter's own timing over one 30-minute window, read .
What each seller accepts
13 parameters · 4 sellers- include_reasoningevery seller
- logprobsevery seller
- max_tokensevery seller
- reasoningevery seller
- reasoning_effortevery seller
- response_formatevery seller
- seedevery seller
- structured_outputsevery seller
- temperatureevery seller
- tool_choiceevery seller
- toolsevery seller
- top_logprobsevery seller
- top_pevery seller
All 13 are accepted by every one of the 4 sellers that publish a list. Each list is the seller's own declaration in OpenRouter's endpoints API, read , and nothing here tested one.
Price position
342 priced modelsBlended $2.40 per million tokens, 3 in to 1 out: cheaper than 94 of the 342 priced models in the catalog. Each bar is one stretch of the catalog on a logarithmic price scale; the mark is this listing.
Measured quality
145 models with an index · 132 on a design arena boardEach mark is one board this model stands on, placed at the share of that board it scores above, and led to its number in the list under the rule. The columns and the ledger below carry the figures.
- Agentic game development · Agents1155 Elowin rate 47.2% · source rank 10the board, 25 ranked
- Android native · Agents1279 Elowin rate 61.0% · source rank 3the board, 41 ranked
- blender3d · Agents1121 Elowin rate 43.7% · source rank 11the board, 25 ranked
- codetovideo · Agents1166 Elowin rate 49.2% · source rank 17the board, 26 ranked
- Full stack · Agents1237 Elowin rate 54.7% · source rank 10the board, 44 ranked
- Godot game development · Agents1247 Elowin rate 53.3% · source rank 4the board, 34 ranked
- HTML slides · Agents1146 Elowin rate 60.5% · source rank 6the board, 25 ranked
- Mobile apps · Agents1224 Elowin rate 55.0% · source rank 15the board, 49 ranked
- pdf · Agents1231 Elowin rate 57.6% · source rank 13the board, 36 ranked
- python-pptx slides · Agents1214 Elowin rate 50.3% · source rank 11the board, 29 ranked
- spreadsheet · Agents1277 Elowin rate 48.0% · source rank 7the board, 32 ranked
- Web apps · Agents1230 Elowin rate 55.2% · source rank 15the board, 47 ranked
- 3D scenes · Models1254 Elowin rate 51.8% · source rank 27the board, 113 ranked
- ASCII art · Models1256 Elowin rate 56.6% · source rank 15the board, 67 ranked
- Code · Models1278 Elowin rate 51.9% · source rank 24the board, 120 ranked
- Data visualisation · Models1281 Elowin rate 51.5% · source rank 18the board, 110 ranked
- Game development · Models1280 Elowin rate 52.9% · source rank 24the board, 120 ranked
- SVG · Models1221 Elowin rate 47.8% · source rank 25the board, 83 ranked
- UI components · Models1283 Elowin rate 52.5% · source rank 26the board, 116 ranked
- Websites · Models1276 Elowin rate 51.4% · source rank 27the board, 126 ranked
Position 30 of the 258 models Epoch has placed, which is not the population of either board above, on a scale anchored at Claude 3.5 Sonnet 130 and GPT-5 150. Epoch states that absolute values are meaningless by themselves and that comparisons between models are what the index is for, that values are not linearly related to benchmark accuracy, and that a highly specialised model may score low while being capable in its own domain.
Benchmark results
2 of 6 boards · 37 results| Task | All runs | Effort | Score |
|---|---|---|---|
| Chess Puzzles | 167 | xhigh | 0.33 |
| high | 0.33 | ||
| EBR-bench | 34 | xhigh | 0.30 |
| FrontierMath-Tier-4-v2-Private | 84 | xhigh | 0.32 |
| FrontierMath-Tiers-1-3-v2-Private | 128 | xhigh | 0.66 |
| Furniture Assembly | 42 | xhigh | 0.40 |
| GPQA diamond | 342 | xhigh | 0.93 |
| high | 0.94 | ||
| MirrorCode | 48 | xhigh | 0.41 |
| Mystery Game Puzzles | 149 | xhigh | 0.34 |
| OTIS Mock AIME 2024-2025 | 319 | xhigh | 0.99 |
| high | 0.98 | ||
| SimpleQA Verified | 106 | xhigh | 0.49 |
| high | 0.49 |
Mean accuracy, 0 to 1, on the task the row names, as Epoch AI runs it and averaged over repeated runs. All eighteen of their tasks, including the held-out FrontierMath tiers, named as they name them. The model id carries the effort setting the run used. Epoch AI Benchmarking Hub names these runs grok-4.6_xhigh and grok-4.6_high, and the effort beside each score says which run it is. A run missing from a task's rows has no published result there, which is not a score of zero. All runs is how many runs the publisher scored on the task, every model's. Read from epoch.ai ↗, which dates its own data . Epoch AI Benchmarking Hub data, by Epoch AI, under CC BY 4.0.
| Task | Score |
|---|---|
| AMPS_Hard | 97 |
| code_completion | 76.09 |
| code_generation | 77.47 |
| connections | 100 |
| consecutive_events | 78.15 |
| integrals_with_game | 85 |
| javascript | 72.73 |
| logic_with_navigation | 82 |
| math_comp | 97.06 |
| olympiad | 91.21 |
| paraphrase | 71.78 |
| plot_unscrambling | 67.09 |
| python | 55 |
| simplify | 69.33 |
| spatial | 98 |
| story_generation | 78.27 |
| summarize | 68.08 |
| tablejoin | 43.42 |
| tablereformat | 100 |
| theory_of_mind | 86.54 |
| typescript | 43.33 |
| typos | 84 |
| zebra_puzzle | 95.50 |
Percent correct on the task the row names, higher is better. All 23 published columns; there is no average to carry, because the site computes its headline in the browser from a weighting this does not reimplement. LiveBench names this run grok-4.6, with no effort setting written on it. Read from livebench.ai ↗, which dates its own data . LiveBench, Apache-2.0. Results produced by the LiveBench maintainers on questions refreshed monthly.
2 of the 6 openly licensed boards Vitna mirrors carry a result for Grok 4.6 (batch), each printed as its publisher published it and none added to another. MTEB, HELM, Vectara hallucination leaderboard and Aider polyglot have not published one under a name that matches its listing, and names are matched to the catalog exactly and never guessed. A task that is a link opens its column on /capability, with every run scored on it; an effort that is a link opens its run's row there, where the run is ranked on Vitna's scale. Every board, with its licence and what its publisher measures, is on /benchmarks.
Source: OpenRouter (openrouter.ai/models), as of 17 Nov 16:29 UTC. sha256 0bb53a77713f over 853,270 bytes.
Catalog: OpenRouter's public model list.
Intelligence, coding and agentic indexes: Artificial Analysis, as published by OpenRouter.
Elo, win rate and rank: Design Arena, as published by OpenRouter.
Source: the Epoch AI Capabilities Index (epoch.ai/eci), as of 20 Nov 2026. 258 models placed over 2917 benchmark results.
Epoch AI Benchmarking Hub data, by Epoch AI, under CC BY 4.0.
Vitna ran none of these evaluations and adjusted none of these figures.
