Grok 4.6
xAI's frontier model for coding, agentic tasks, and knowledge work, positioned around long-running agents with enhanced self-testing, launched in Cursor, Grok Build, and the xAI API.
Grok 4.6 is available on MiniRouter as . The model page has current pricing, provider routes and API setup.
About the model
- Context: 500K. Per xAI's docs; prompts above 200K tokens bill at a higher rate
- Long-prompt price: $4 / $12. Per 1M input / output tokens above 200K prompt tokens
- Reasoning effort: low – xhigh. Four publisher-described effort levels; high is the default
- Weights: Not mentioned. The announcement does not include a weights release
Publisher pricing
Pricing below comes from the linked publisher sources and does not represent a MiniRouter route or price.
| Phase | Input / 1M | Output / 1M | Blended / 1M | Effective | Source |
|---|---|---|---|---|---|
| Standard | $2.00 | $6.00 | $3.00 | Aug 12, 2026 – open-ended | xAI ↗ |
Publisher-reported benchmarks
Results from the publisher’s announcement, grouped by unit.
| Benchmark | Reported result | Reported by | As of | Source |
|---|---|---|---|---|
| DeepSWE v1.1 | 65.9% | xAI | Aug 12, 2026 | Announcement ↗ |
| FrontierCode v1.1 (Extended) | 61.3% | xAI | Aug 12, 2026 | Announcement ↗ |
| CursorBench v3.2 | 69.9% | xAI | Aug 12, 2026 | Announcement ↗ |
| APEX-SWE | 56.4% | xAI | Aug 12, 2026 | Announcement ↗ |
| Terminal-Bench v3.0 | 26% | xAI | Aug 12, 2026 | Announcement ↗ |
| Terminal-Bench v2.1 | 82.4% | xAI | Aug 12, 2026 | Announcement ↗ |
| GPQA Diamond | 89.4% | xAI | Aug 12, 2026 | Announcement ↗ |
| Benchmark | Reported result | Reported by | As of | Source |
|---|---|---|---|---|
| GDPVal-AA v2 | 1753 | xAI | Aug 12, 2026 | Announcement ↗ |
xAI sources competitor benchmark figures from those vendors' published system cards and leaderboards.
xAI's docs list cached input at $0.50 per 1M tokens and a fast variant at twice the base price.
xAI reports the benchmark figures; MiniRouter has not reproduced them.
Independent measurements
Artificial Analysis evaluations of the first-party API.
Grok 4.6 (high) ranks #10 of 462 model families on their Intelligence Index. Their headline uses the highest-scoring effort setting; every listed setting is below.
- Intelligence Index
- 44.3
- Composite of their evaluations
- Coding Index
- 76.8
- Coding evaluations only
- Blended price
- $3
- USD per 1M tokens, 3:1 input to output, first-party list
- Output speed
- 70
- Median tokens per second
- First answer token
- 39s
- Median seconds, including reasoning
| Setting | Intelligence Index | Coding Index | Blended price | Output speed | First answer token |
|---|---|---|---|---|---|
| Grok 4.6 (high) ↗ | 44.3 | 76.8 | $3 | 70 | 39s |
| Grok 4.6 (xhigh) ↗ | 44.2 | 75.9 | $3 | 69 | 43s |
| Grok 4.6 (medium) ↗ | 42.8 | 74.4 | $3 | 68 | 26s |
| Grok 4.6 (low) ↗ | 35.1 | 66.3 | $3 | 55 | 3.1s |
Source: Artificial Analysis ↗, read Sep 21, 2026. Measured on the model's first-party API, not a MiniRouter route.
Keep exploring
Compare published results in the model rankings and coding benchmarks, or read more model releases.
For MiniRouter product updates, visit the changelog.
Sources
- Read at xAI ↗
xAI announcement
Published Aug 12, 2026
- Read at Artificial Analysis ↗
Artificial Analysis model page
Published Sep 21, 2026
Editorial reference revised 2026-09-29. Catalog status comes from the generated routing snapshot.