GPT OSS 120BvsMercury 2
GPT OSS 120B is cheaper than Mercury 2 at $0.18/M vs $0.25/M input tokens.
| Spec | GPT OSS 120B | Mercury 2 |
|---|---|---|
| Input price | $0.18/M tokens | $0.25/M tokens |
| Output price | $0.8/M tokens | $0.75/M tokens |
| Context window | 131K tokens | 128K tokens |
| Weights | Open | — |
| Free API (OpenRouter) | No | No |
| Released | Aug 2025 | Mar 2026 |
Our Verdict
GPT OSS 120B
Mercury 2Runner-up
Not enough votes to call it. On the specs, GPT OSS 120B has the edge: bigger model tier, major provider backing.
Too close to call
Reviewing agent-written code?See a Brief PR report
API pricing
Cost per 1M tokens
GPT OSS 120B
Input
$0.18
1.4× cheaper
Output
$0.80
Mercury 2
Input
$0.25
Output
$0.75
1.1× cheaper
GPT OSS 120B wins input (1.4× cheaper)·Mercury 2 wins output (1.1× cheaper)
Where to run it
20 hosts, cheapest first
GPT OSS 120B19 hosts
AkashMLbf16$0.03 in$0.17 out131k100% upCoreWeavefp4$0.03 in$0.17 out131k100% upDekaLLMbf16$0.03 in$0.18 out131k99.1% upDeepInfrabf16$0.04 in$0.17 out131k99.7% upNovitafp4$0.05 in$0.25 out131k95.5% upDigitalOcean$0.06 in$0.39 out128k100% up
13 more hostsFewer hosts
Mancerfp8$0.06 in$0.50 out131k98% upGoogle Vertex AI$0.09 in$0.36 out131k99.9% upBasetenfp4$0.10 in$0.50 out128k100% upParasailfp4$0.10 in$0.75 out131k100% upSambaNova$0.14 in$0.95 out131k99% upAmazon Bedrock$0.15 in$0.60 out131k100% upGroq$0.15 in$0.60 out131k100% upNebiusfp4$0.15 in$0.60 out131k99.8% upPhala$0.15 in$0.60 out131k96.1% upSiliconFlowfp8$0.15 in$0.60 out131k95.3% upCerebrasfp16$0.35 in$0.75 out131k100% upTogetherdegraded$0.15 in$0.60 out131k97.7% upMaradegraded$0.15 in$0.75 out131k82% up
Mercury 21 host
Per million tokens. Prices and uptime via OpenRouter, checked 13 Sep 2026.
Writing DNA
Style Comparison
Similarity
94%
GPT OSS 120B uses 1.9x more emoji
GPT OSS 120B
Mercury 2
52%Vocabulary53%
19wSentence Length22w
0.28Hedging0.40
7.4Bold6.6
1.8Lists2.6
0.15Emoji0.08
0.73Headings0.87
0.17Transitions0.14
Based on 21 + 23 text responses
Research
What we learned reading every model
FAQ
Common questions
Keep exploring
