xAI's frontier model for long-running agents, coding, knowledge work, and visual projects
Specification
xai listing
Context
500K
Max output
500K
Released
2026-08-12
Knowledge cutoff
2026-02-01
Retires
—
Open weights
no
Input
text, image, pdf
Output
text
Price
US dollars per million tokens · xai list price
Input
$2.00
Output
$6.00
Cache read
$0.50
Cache write
$2.00
Long-context price tiers
the rate above each context threshold
| Above | Price | $/M |
|---|---|---|
| 200K | Cache read | $1.00 |
| 200K | Input | $4.00 |
| 200K | Output | $12.00 |
Quality
9 benchmarks · 2 sources
Composite
87th
percentile of 334 scored models
Rank
45 / 334
±0.07 sd
Evidence
9 × 2
3 or more benchmarks, from 2 or more sources
Effort range
1.1
points between effort settings
reasoning92nd4/4 bench
math86th3/7 bench
knowledge69th1/1 bench
general79th1/1 bench
By reasoning effort
every setting placed on the same scale as the leaderboard
| Setting | Composite | Benchmarks behind it |
|---|---|---|
| high | 93rd | 4 |
| xhigh | 87th | 8 |
Every score
one row per source and configuration — nothing averaged away
| Benchmark | Score | Configuration | Source | Run |
|---|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 99.2 ±0.5 | xhigh | Epoch AI | 2026-08-14 |
| ↳ | 97.8 ±1.0 | high | Epoch AI | 2026-08-12 |
| Settings differ by 1.4 points on this benchmark. | ||||
| GPQA diamond | 94.0 ±1.4 | high | Epoch AI | 2026-08-12 |
| ↳ | 93.2 ±1.5 | xhigh | Epoch AI | 2026-08-14 |
| Settings differ by 0.8 points on this benchmark. | ||||
| Chess Puzzles | 40.0 ±4.9 | high | Epoch AI | 2026-08-12 |
| ↳ | 31.0 ±4.6 | xhigh | Epoch AI | 2026-08-14 |
| Settings differ by 9.0 points on this benchmark. | ||||
| Mystery Game Puzzles | 34.0 ±4.8 | xhigh | Epoch AI | 2026-08-14 |
| FrontierMath-Tiers-1-3-v2 | 66.0 ±2.8 | xhigh | Epoch AI | 2026-08-14 |
| LiveBench | 79.0 | default · 23 runs | LiveBench | 2026-06-25 |
| SimpleQA Verified | 49.3 ±1.6 | high | Epoch AI | 2026-08-27 |
| ↳ | 48.9 ±1.6 | xhigh | Epoch AI | 2026-08-27 |
| Settings differ by 0.4 points on this benchmark. | ||||
| EBR-bench | 30.5 ±1.6 | xhigh | Epoch AI | 2026-08-18 |
| FrontierMath-Tier-4-v2 | 31.7 ±7.4 | xhigh | Epoch AI | 2026-08-14 |
The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.
Available from 40 hosts
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| xaivendor | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| kenari | free | free | — | — | 500K | 500K | — |
| 302ai | $2.00 | $6.00 | — | — | 500K | 500K | — |
| abacus | $2.00 | $6.00 | $0.50 | — | 500K | 33K | — |
| aihubmix | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| amazon-bedrock · global | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| azure_ai | $2.00 | $6.00 | $0.50 | — | 200K | 128K | — |
| bedrock_converse · global | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| cloudflare-ai-gateway | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| crossmodel · x-ai | $2.00 | $6.00 | $0.50 | $2.00 | 500K | 500K | — |
| edenai | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| github-copilot | $2.00 | $6.00 | $0.50 | — | 500K | 128K | — |
| google-vertex | $2.00 | $6.00 | $0.50 | — | 524K | 500K | — |
| kilo · x-ai | $2.00 | $6.00 | $0.50 | — | 500K | 450K | — |
| llmgateway | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| llmgateway-providers | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| llmgateway-providers · aws-bedrock | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| llmgateway-providers · vertex-openai | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| merge-gateway | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| nano-gpt · x-ai | $2.00 | $6.00 | $0.50 | — | 500K | 450K | — |
| ofox · x-ai | $2.00 | $6.00 | $0.50 | — | 500K | 66K | — |
| opencode | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| opencode-go | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| openrouter | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| openrouter · x-ai | $2.00 | $6.00 | $0.50 | — | 500K | 450K | — |
| opper | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| orcarouter · grok | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| perplexity | $2.00 | $6.00 | $0.50 | — | — | — | — |
| perplexity-agent | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| requesty | $2.00 | $6.00 | $0.50 | $2.00 | 500K | 500K | — |
| vercel · spacexai | $2.00 | $6.00 | $0.50 | — | 500K | 500K | — |
| vertex_ai | $2.00 | $6.00 | $0.50 | — | 524K | 524K | — |
| amazon-bedrock | $2.20 | $6.60 | $0.55 | — | 500K | 500K | — |
| amazon-bedrock · us | $2.20 | $6.60 | $0.55 | — | 500K | 500K | — |
| bedrock_converse · us | $2.20 | $6.60 | $0.55 | — | 500K | 500K | — |
| bedrock_mantle · bedrock-mantle | $2.20 $2.64 | $6.60 $7.92 | $0.55 $0.66 | — | 500K | 500K | — |
| venice | $2.27 | $6.80 | $0.57 | — | 500K | 200K | — |
| databricks | $2.50 | $7.50 | $0.625 | $2.50 | 500K | — | — |
| azure | — | — | — | — | 200K | 128K | — |
| neon | — | — | — | — | 500K | 524K | — |
Price history
append-only observations · a listing writes a row only when its price moves
| Date | Host | Field | $/M |
|---|---|---|---|
| 2026-09-02 | bedrock_mantle · bedrock-mantle | Cache read | $0.55 |
| 2026-09-09 | bedrock_mantle · bedrock-mantle | Cache read | $0.55 |
| 2026-09-02 | bedrock_mantle · bedrock-mantle | Input | $2.20 |
| 2026-09-09 | bedrock_mantle · bedrock-mantle | Input | $2.20 |
| 2026-09-02 | bedrock_mantle · bedrock-mantle | Output | $6.60 |
| 2026-09-09 | bedrock_mantle · bedrock-mantle | Output | $6.60 |
More from xAI
most-hosted first