DeepSeek V4.1 Flash vs Grok 4.6
Published test results, prices, limits and features, side by side. Share the link to show someone exactly this comparison.
DeepSeek V4.1 FlashDeepSeek
Grok 4.6xAI
At a glance
Capability index
- DeepSeek V4.1 Flash155.6
- Grok 4.6156.3
Grok 4.6 is 0.7 points higher.
Price per 1M tokens
- DeepSeek V4.1 Flash$0.464
- Grok 4.6$3.00
DeepSeek V4.1 Flash is 6.5 times cheaper.
Context window
- DeepSeek V4.1 Flash1M tokens
- Grok 4.6500K tokens
DeepSeek V4.1 Flash takes in 2.1 times as much.
Released
- DeepSeek V4.1 FlashSep 10, 2026
- Grok 4.6Aug 12, 2026
DeepSeek V4.1 Flash is 29 days newer.
Where each scores higher
On the 21 tests both have taken, each on its own scale. A gap of a point or two can sit within a test’s margin of error.
DeepSeek V4.1 Flash
- LiveBench overall+3.1 points
- Arena text+20 points
- LiveBench coding+3.3 points
- Arena WebDev+1 points
- Arena coding+16 points
- LiveBench mathematics+0.7 points
- FrontierMath Tiers 1–3+1.4 points
- Arena math+48 points
- LiveBench agentic coding+20.2 points
- Arena instruction following+25 points
- LiveBench data analysis+5.4 points
- Chess puzzles+15.0 points
- Arena hard prompts+18 points
Grok 4.6
- Mock AIME 2024–2025+0.8 points
- FrontierMath Tier 4+4.9 points
- SimpleQA Verified+6.4 points
- LiveBench language+2.5 points
- LiveBench instructions+1.8 points
- Arena creative writing+4 points
- LiveBench reasoning+3.8 points
- GPQA Diamond+4.2 points
2 more tests have results for only one of them; the table below lists every result.
Everything side by side
| Measure | DeepSeek V4.1 FlashDeepSeek | Grok 4.6xAI |
|---|---|---|
| Model | ||
| Maker | DeepSeek | xAI |
| Sold here by | engy | xAI |
| API model ID | deepseek-v4.1-flash | grok-4.6 |
| Released | Sep 10, 2026 | Aug 12, 2026 |
| Knowledge cutoff | May 2025 | Feb 1, 2026 |
| Weights | Open: downloadable | Closed: hosted access only |
| Developer’s country | China | United States |
| Hosts selling it | 80 | 37including xAI directly |
| Training compute | 4.3 × 10²⁴ FLOPEpoch AI estimate | Not published |
| Test results | ||
| Capability index | 155.6#31 of 277 | 156.3 (best of these models)#27 of 277 |
| LiveBench overall | 81.1% (best of these models)#9 of 70 | 78.0%#19 of 70 |
| Arena text | 1475 (best of these models)#38 of 415 | 1455#70 of 415 |
| LiveBench coding | 80.0% (best of these models)#23 of 70 | 76.8%#45 of 70 |
| Arena WebDev | 1619 (best of these models)#24 of 142 | 1617#25 of 142 |
| Arena coding | 1526 (best of these models)#27 of 410 | 1510#60 of 410 |
| LiveBench mathematics | 93.3% (best of these models)#23 of 70 | 92.6%#28 of 70 |
| Mock AIME 2024–2025 | 98.3%#24 of 303 | 99.2% (best of these models)#17 of 303 |
| FrontierMath Tiers 1–3 | 67.4% (best of these models)#30 of 117 | 66.0%#31 of 117 |
| FrontierMath Tier 4 | 26.8%#41 of 73 | 31.7% (best of these models)#35 of 73 |
| Arena math | 1495 (best of these models)#22 of 400 | 1447#83 of 400 |
| SimpleQA Verified | 42.9%#43 of 90 | 49.3% (best of these models)#28 of 90 |
| LiveBench agentic coding | 77.3% (best of these models)#1 of 70 | 57.0%#25 of 70 |
| DeepSWE | Not tested | 67.5%#20 of 69 |
| CursorBench | Not tested | 41.4%#31 of 71 |
| LiveBench language | 81.2%#29 of 70 | 83.7% (best of these models)#19 of 70 |
| LiveBench instructions | 70.0%#32 of 70 | 71.9% (best of these models)#21 of 70 |
| Arena creative writing | 1439#61 of 413 | 1443 (best of these models)#55 of 413 |
| Arena instruction following | 1477 (best of these models)#24 of 415 | 1452#63 of 415 |
| LiveBench data analysis | 79.3% (best of these models)#18 of 70 | 73.9%#42 of 70 |
| LiveBench reasoning | 86.7%#32 of 70 | 90.5% (best of these models)#13 of 70 |
| GPQA Diamond | 89.8%#50 of 323 | 94.0% (best of these models)#12 of 323 |
| Chess puzzles | 48.0% (best of these models)#20 of 243 | 33.0%#63 of 243 |
| Arena hard prompts | 1499 (best of these models)#36 of 415 | 1481#64 of 415 |
| Price per million tokens | ||
| Input | $0.04 (best of these models) | $2.00 |
| Output | $0.08 (best of these models) | $6.00 |
| Blended, 3 input to 1 output | $0.05 (best of these models) | $3.00 |
| Cached input | $0.008 (best of these models) | $0.50 |
| Long requests | No separate price reported | $4.00 in, $12.00 outabove 200K tokens |
| The maker’s own price | Not sold directly | This listing |
| Limits | ||
| Context window | 327.7K tokens | 500K tokens (best of these models) |
| Max input | 1M tokens (best of these models) | 372K tokens |
| Max output | 65.5K tokens | 500K tokens (best of these models) |
| Features | ||
| Reads | images and text | images, PDFs and text |
| Produces | text | text |
| Reasoning | Yeseffort none, low, high, max; can be switched off; thinking budget can be set | Yeseffort low, medium, high, xhigh |
| Tool calling | Yes | Yes |
| Structured output | Yes, such as JSON | Yes, such as JSON |
| File attachments | Yes | Yes |
| Temperature setting | Supported | Supported |
Bold marks the better value in each row. Prices are each model’s own list price, or the middle price across hosts when the maker does not sell it, blended as three input tokens for every output token. Test results as published by LiveBench (Apache-2.0), Epoch AI (CC BY 4.0), Datacurve DeepSWE, compiled by Epoch AI (CC BY 4.0), Cursor CursorBench, compiled by Epoch AI (CC BY 4.0), Arena (CC BY 4.0); model details and prices from models.dev (MIT). Confirm pricing with the provider before use.