DeepSeek-V4-Pro-0813 vs gnm-v3: Benchmarks, Pricing & Capabilities Compared
TL;DR — DeepSeek-V4-Pro-0813 wins for general use · gnm-v3 wins for general use.
DeepSeek-V4-Pro-0813 deepseek-ai
- Released
- 2026-08-13
gnm-v3 google
- Released
- 2026-09-01
Benchmark comparison
No benchmarks reported on both models.
Pricing comparison
| Metric | DeepSeek-V4-Pro-0813 | gnm-v3 |
|---|---|---|
| Input ($/Mtok) | — | — |
| Output ($/Mtok) | — | — |
| Cached input ($/Mtok) | — | — |
| Cost per 1M-token roundtrip (1M in + 1M out) | — | — |
Context window & modalities
| Attribute | DeepSeek-V4-Pro-0813 | gnm-v3 |
|---|---|---|
| Context window | — | — |
| Input modalities | ||
| Output modalities | ||
| Knowledge cutoff | — | — |
Verdict by use case
Coding
Insufficient data
Basis: SWE-bench
No shared coding benchmark.
Reasoning
Insufficient data
Basis: GPQA / MMLU
No shared reasoning benchmark.
Math
Insufficient data
Basis: MATH / AIME
No shared math benchmark.
Long context
Insufficient data
Basis: Context window
Context window not reported.
Cost
Insufficient data
Basis: Input $/Mtok
Input pricing not reported on both.
Changelog & releases
DeepSeek-V4-Pro-0813
Released 2026-08-13
gnm-v3
Released 2026-09-01