Gemini 4 Argon
Category scores
Scores per capability area with the model's rank among ranked models in that category.
- 1Math92.8
- 2Reasoning#386.9
- 3Agentic & tool use79.6
- 4Knowledge74.9
- 5Coding#674.0
Benchmark results (25)
As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.
Agentic & tool use
| MazeBench | 12 | #5 of 42 |
| AutomationBench | 51.3 | #7 of 25 |
| OSWorld 2.0 | 69.2 | #3 of 24 |
| Terminal-Bench 4.0 | 57.40 | #5 of 18 |
| Agents' Last Exam | 39.5 | #6 of 17 |
| CWE-bench v1 | 68.0 | #1 of 11 |
| Terminal-Bench-Science 0.1 (6x verifier timeout) | 57.6 | #1 of 1 |
Reasoning
| ARC-AGI-2 | 91.3 | #5 of 41 |
| ARC-AGI-1 | 98.50 | #1 of 36 |
| ARC-AGI-3 | 46.3 | #3 of 22 |
| STEMBench | 2.44 | #8 of 9 |
| ARC-AGI-3 Provider Adapter | 97.30 | #2 of 8 |
| GraphWalks BFS 256K–1M | 84.2 | #1 of 4 |
| Graphwalks BFS 128K | 99.7 | #1 of 2 |
Coding
| DeepSWE | 77.9 | #3 of 50 |
| Bug Hunt Bench | 40.7 | #8 of 23 |
| FrontierSWE v2 | 55.0 | #5 of 20 |
| PostTrainBench v1.1 | 45.3 | #2 of 15 |
Multimodal
| Design Arena Website | 1315 | #8 of 91 |
| LVBench | 91.7 | #1 of 6 |
| Chartography (no tools) | 71.6 | #1 of 5 |
Math
| FrontierMath v2 (Tiers 1-3) | 85.965 | #10 of 64 |
| FrontierMath v2 (Tier 4) | 78.049 | #9 of 51 |
Knowledge
| LABBench2 | 88.8 | #1 of 7 |
Instruction following
| Gray Swan IPI (15 attempts)(lower is better) | 0.7 | #1 of 12 |
Pricing by source
USD per 1M tokens.
| Source | Input | Output | Cached in |
|---|---|---|---|
| ✓ Official (Google) checked Oct 17, 2026 · gemini-4-argon | $2.00 | $10.00 | — |
| BenchLM list price | $2.00 | $10.00 | $0.10 |
| OpenRouter google/gemini-4-argon | $2.00 | $10.00 | $0.10 |
Paid tier, prompts up to 200K tokens. Several prices change on Dec 1, 2026 / Jan 1, 2027.
Google's Gemini API pricing page lists gemini-4-argon on the paid tier, not the Free tier. Standard input/output is $2.00 / $10.00 per million tokens through November 30, 2026, then $4.00 / $20.00 starting December 1; implicit cached input is $0.10 / $0.20 across those dates, and explicit caching is unsupported. Batch and Flex input/output is $1.00 / $5.00 through November 30, then $2.00 / $10.00 from December 1, with implicit cached input at $0.05 / $0.10. Priority input/output is $3.60 / $18.00 through November 30, then $7.20 / $36.00 from December 1, with implicit cached input at $0.18 / $0.36.
Arena ratings
Human-preference ratings from LMArena. Best-rated variant shown.
| Text(High) | 1532 | #1 · 4.9K votes |
| WebDev (coding)(High) | 1678 | #10 · 1.7K votes |
Speed
Measured by us through OpenRouter (median of trials).
Not measured yet.
Related models
Same family
- Gemini 3.1 Pro64.7
- Gemini 3.8 Flash69.2
Similar score
- Claude Fable 5.183.2 · $20.00
- Claude Sonnet 5.582.2 · $4.00
- GPT-6 Astra84.0 · $20.00
- GPT-6.1 Sol84.2 · $4.00
- Claude Fable 581.3 · $20.00