DeepSeek V4.1 Flash vs Gemini 4 Argon
Published test results, prices, limits and features, side by side. Share the link to show someone exactly this comparison.
DeepSeek V4.1 FlashDeepSeek
Gemini 4 ArgonGoogle
At a glance
Capability index
- DeepSeek V4.1 Flash155.6
- Gemini 4 Argon164.6
Gemini 4 Argon is 9.0 points higher.
Price per 1M tokens
- DeepSeek V4.1 Flash$0.464
- Gemini 4 Argon$4.00
DeepSeek V4.1 Flash is 8.6 times cheaper.
Context window
- DeepSeek V4.1 Flash1M tokens
- Gemini 4 Argon1M tokens
They take in about the same.
Released
- DeepSeek V4.1 FlashSep 10, 2026
- Gemini 4 ArgonOct 13, 2026
Gemini 4 Argon is 33 days newer.
Where each scores higher
On the 13 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
No shared test where it scores higher.
Gemini 4 Argon
- Arena text+50 points
- Arena WebDev+60 points
- Arena coding+35 points
- Mock AIME 2024–2025+1.7 points
- FrontierMath Tiers 1–3+18.6 points
- FrontierMath Tier 4+51.2 points
- Arena math+33 points
- SimpleQA Verified+33.1 points
- Arena creative writing+77 points
- Arena instruction following+52 points
- GPQA Diamond+6.6 points
- Chess puzzles+11.0 points
- Arena hard prompts+51 points
9 more tests have results for only one of them; the table below lists every result.
Everything side by side
| Measure | DeepSeek V4.1 FlashDeepSeek | Gemini 4 ArgonGoogle |
|---|---|---|
| Model | ||
| Maker | DeepSeek | |
| Sold here by | engy | |
| API model ID | deepseek-v4.1-flash | gemini-4-argon |
| Released | Sep 10, 2026 | Oct 13, 2026 |
| Knowledge cutoff | May 2025 | Mar 2026 |
| Weights | Open: downloadable | Closed: hosted access only |
| Developer’s country | China | United States |
| Hosts selling it | 80 | 24including Google directly |
| Training compute | 4.3 × 10²⁴ FLOPEpoch AI estimate | Not published |
| Test results | ||
| Capability index | 155.6#31 of 277 | 164.6 (best of these models)#5 of 277 |
| LiveBench overall | 81.1%#9 of 70 | Not tested |
| Arena text | 1475#38 of 415 | 1525 (best of these models)#1 of 415 |
| LiveBench coding | 80.0%#23 of 70 | Not tested |
| Arena WebDev | 1619#24 of 142 | 1678 (best of these models)#9 of 142 |
| Arena coding | 1526#27 of 410 | 1561 (best of these models)#1 of 410 |
| LiveBench mathematics | 93.3%#23 of 70 | Not tested |
| Mock AIME 2024–2025 | 98.3%#24 of 303 | 100.0% (best of these models)#1 of 303 |
| FrontierMath Tiers 1–3 | 67.4%#30 of 117 | 86.0% (best of these models)#10 of 117 |
| FrontierMath Tier 4 | 26.8%#41 of 73 | 78.0% (best of these models)#15 of 73 |
| Arena math | 1495#22 of 400 | 1529 (best of these models)#1 of 400 |
| SimpleQA Verified | 42.9%#43 of 90 | 76.0% (best of these models)#1 of 90 |
| LiveBench agentic coding | 77.3%#1 of 70 | Not tested |
| CursorBench | Not tested | 52.9%#6 of 71 |
| LiveBench language | 81.2%#29 of 70 | Not tested |
| LiveBench instructions | 70.0%#32 of 70 | Not tested |
| Arena creative writing | 1439#61 of 413 | 1516 (best of these models)#2 of 413 |
| Arena instruction following | 1477#24 of 415 | 1528 (best of these models)#1 of 415 |
| LiveBench data analysis | 79.3%#18 of 70 | Not tested |
| LiveBench reasoning | 86.7%#32 of 70 | Not tested |
| GPQA Diamond | 89.8%#50 of 323 | 96.4% (best of these models)#1 of 323 |
| Chess puzzles | 48.0%#20 of 243 | 59.0% (best of these models)#9 of 243 |
| Arena hard prompts | 1499#36 of 415 | 1550 (best of these models)#1 of 415 |
| Price per million tokens | ||
| Input | $0.04 (best of these models) | $2.00 |
| Output | $0.08 (best of these models) | $10.00 |
| Blended, 3 input to 1 output | $0.05 (best of these models) | $4.00 |
| Cached input | $0.008 (best of these models) | $0.10 |
| Audio input | Not reported | $2.00 |
| The maker’s own price | Not sold directly | This listing |
| Limits | ||
| Context window | 327.7K tokens | 1M tokens (best of these models) |
| Max input | 1M tokens | 1M tokens |
| Max output | 65.5K tokens (best of these models) | 32.8K tokens |
| Features | ||
| Reads | images and text | audio, images, PDFs, text and video |
| Produces | text | text |
| Reasoning | Yeseffort none, low, high, max; can be switched off; thinking budget can be set | Yeseffort low, medium, high |
| 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), 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.