GPT OSS 120B

Open GPT reasoning model for self-hosted agents and controllable deployments

Strongest at
Math
Top 22% of the past year's models · one benchmark so far
Price per 1M tokens
$0.15 / $0.6
input / output · OpenRouter
Context window
131.1K
tokens in · 32.8K out

Capability profile

Scored against 91 models released in the past year. 50 is the average model; further out is better.

This modelMedian of the past year
5090Median of the past yearCoding: 47 · Better than 27%Writing: 29 · Better than 7%Reasoning: 29 · Better than 5%Math: 64 · Top 22%Science: 54 · Better than 19%Computer use: 31 · Better than 0%CodingWritingReasoningMathScienceComputer use
Coding47Better than 27% · 1 test
Writing29Better than 7%
Reasoning29Better than 5%
Math64Top 22% · 1 test
Science54Better than 19%
Computer use31Better than 0%

No results yet: Reading & documents, Visual reasoning

Where these scores come from

Each capability combines the published benchmarks listed here, rescaled so 50 is the reference average and 15 points is one standard deviation. A hollow dot rests on a single benchmark.

Price against capability

The shaded corner holds anything better and cheaper. Prices blend input and output 3 to 1.

1 model from the past year scores higher on Math for less: Claude Haiku 5.5.

$0.1$0.3$1$3$10$3020406080GPT-6.1 Terra: 76, $2 per 1MDeepSeek V4.1 Flash 1017: 63, $0.52 per 1MMuse Spark Open: 60, $2 per 1MClaude Haiku 5.5: 66, $0.2 per 1MMistral Large 4: 54, $2.06 per 1MGemini 4 Argon: 75, $4 per 1MGPT-6.1 Sol: 78, $4 per 1MClaude Sonnet 5.5: 75, $4 per 1MGPT-6 Sol: 74, $4 per 1MGPT-6 Luna: 63, $0.2 per 1MClaude Opus 5.5: 77, $8 per 1MGrok 4.7: 55, $3 per 1MDeepSeek V4.1 Flash: 60, $0.21 per 1MGPT-6 Astra: 78, $20 per 1MMuse Spark 1.3: 65, $2 per 1MGemini 3.8 Flash: 58, $1.5 per 1MClaude Fable 5.1: 76, $20 per 1MGLM-5.3-Flash: 47, $0.24 per 1MDeepSeek V4 Flash Vision Exp: 48, $0.32 per 1MQwen3.8 27B: 42, $1.12 per 1MGLM-5.3: 56, $2.15 per 1MGemini 3.7 Flash: 62, $1.5 per 1MDeepSeek V4 Pro 0813: 60, $0.99 per 1MGrok 4.6: 60, $3 per 1MMuse Spark 1.2: 56, $2 per 1MQwen3.8 Max: 62, $3 per 1MDeepSeek V4 Flash 0731: 55, $0.15 per 1MClaude Opus 5: 73, $10 per 1MGemini 3.6 Flash: 52, $1.5 per 1MGemini 3.5 Flash Lite: 35, $0.85 per 1MKimi K3: 61, $6 per 1MQwen3.7 Flash: 48, $0.055 per 1MGPT-5.6 Terra: 70, $4.5 per 1MGPT-5.6 Sol: 73, $8 per 1MGPT-5.6 Luna: 63, $0.45 per 1MGrok 4.5: 55, $3 per 1MClaude Sonnet 5: 62, $4 per 1MGLM-5.2: 54, $2.15 per 1MKimi K2.7 Code: 46, $1.71 per 1MClaude Fable 5: 75, $20 per 1MQwen3.7 Plus: 54, $0.7 per 1MClaude Opus 4.8: 67, $10 per 1MQwen3.7 Max: 53, $3.75 per 1MGemini 3.5 Flash: 54, $3.38 per 1MGPT-5.5 Instant: 42, $11.2 per 1MDeepSeek V4 Pro: 49, $1.19 per 1MDeepSeek V4 Flash: 27, $0.18 per 1MGPT-5.5 Pro: 72, $67.5 per 1MGPT-5.5: 68, $11.2 per 1MQwen3.6 27B: 45, $1.35 per 1MKimi K2.6: 49, $1.71 per 1MGrok 4.3: 39, $1.56 per 1MGrok Build 0.1: 23, $1.25 per 1MClaude Opus 4.7: 60, $10 per 1MQwen3.6 35B-A3B: 49, $0.84 per 1MMuse Spark 1.1: 50, $2 per 1MGLM-5.1: 51, $2.15 per 1MGemma 4 31B IT: 58, $0.25 per 1MQwen3.6 Plus: 37, $1.12 per 1MGPT-5.4 nano: 48, $0.46 per 1MGPT-5.4 mini: 44, $1.69 per 1MGPT-5.4 Pro: 67, $67.5 per 1MGPT-5.4: 64, $5.62 per 1MGemini 3.1 Flash Lite: 49, $0.56 per 1MQwen3.5 Flash: 47, $0.18 per 1MQwen3.5 9B: 53, $0.19 per 1MQwen3.5 35B-A3B: 56, $0.69 per 1MGemini 3.1 Pro Preview: 55, $4.5 per 1MClaude Sonnet 4.6: 54, $6 per 1MQwen3.5 Plus: 63, $0.9 per 1MQwen3.5 397B-A17B: 52, $1.35 per 1MGLM-5: 60, $1.55 per 1MClaude Opus 4.6: 57, $10 per 1MGLM-4.7: 48, $1 per 1MGemini 3 Flash Preview: 50, $1.12 per 1MGPT-5.2 Pro: 61, $57.8 per 1MGPT-5.2 Codex: 51, $4.81 per 1MGPT-5.2: 54, $4.81 per 1MDeepSeek V3.2: 64, $0.32 per 1MClaude Opus 4.5: 51, $10 per 1MGPT-5.1 Codex Max: 37, $3.44 per 1MGPT-5.1: 64, $3.44 per 1MGPT OSS 120B: 64, $0.26 per 1MGPT OSS 120BPrice per 1M tokens →↑ Math$0.1$1$1020406080GPT-6.1 Terra: 76, $2 per 1MDeepSeek V4.1 Flash 1017: 63, $0.52 per 1MMuse Spark Open: 60, $2 per 1MClaude Haiku 5.5: 66, $0.2 per 1MMistral Large 4: 54, $2.06 per 1MGemini 4 Argon: 75, $4 per 1MGPT-6.1 Sol: 78, $4 per 1MClaude Sonnet 5.5: 75, $4 per 1MGPT-6 Sol: 74, $4 per 1MGPT-6 Luna: 63, $0.2 per 1MClaude Opus 5.5: 77, $8 per 1MGrok 4.7: 55, $3 per 1MDeepSeek V4.1 Flash: 60, $0.21 per 1MGPT-6 Astra: 78, $20 per 1MMuse Spark 1.3: 65, $2 per 1MGemini 3.8 Flash: 58, $1.5 per 1MClaude Fable 5.1: 76, $20 per 1MGLM-5.3-Flash: 47, $0.24 per 1MDeepSeek V4 Flash Vision Exp: 48, $0.32 per 1MQwen3.8 27B: 42, $1.12 per 1MGLM-5.3: 56, $2.15 per 1MGemini 3.7 Flash: 62, $1.5 per 1MDeepSeek V4 Pro 0813: 60, $0.99 per 1MGrok 4.6: 60, $3 per 1MMuse Spark 1.2: 56, $2 per 1MQwen3.8 Max: 62, $3 per 1MDeepSeek V4 Flash 0731: 55, $0.15 per 1MClaude Opus 5: 73, $10 per 1MGemini 3.6 Flash: 52, $1.5 per 1MGemini 3.5 Flash Lite: 35, $0.85 per 1MKimi K3: 61, $6 per 1MQwen3.7 Flash: 48, $0.055 per 1MGPT-5.6 Terra: 70, $4.5 per 1MGPT-5.6 Sol: 73, $8 per 1MGPT-5.6 Luna: 63, $0.45 per 1MGrok 4.5: 55, $3 per 1MClaude Sonnet 5: 62, $4 per 1MGLM-5.2: 54, $2.15 per 1MKimi K2.7 Code: 46, $1.71 per 1MClaude Fable 5: 75, $20 per 1MQwen3.7 Plus: 54, $0.7 per 1MClaude Opus 4.8: 67, $10 per 1MQwen3.7 Max: 53, $3.75 per 1MGemini 3.5 Flash: 54, $3.38 per 1MGPT-5.5 Instant: 42, $11.2 per 1MDeepSeek V4 Pro: 49, $1.19 per 1MDeepSeek V4 Flash: 27, $0.18 per 1MGPT-5.5 Pro: 72, $67.5 per 1MGPT-5.5: 68, $11.2 per 1MQwen3.6 27B: 45, $1.35 per 1MKimi K2.6: 49, $1.71 per 1MGrok 4.3: 39, $1.56 per 1MGrok Build 0.1: 23, $1.25 per 1MClaude Opus 4.7: 60, $10 per 1MQwen3.6 35B-A3B: 49, $0.84 per 1MMuse Spark 1.1: 50, $2 per 1MGLM-5.1: 51, $2.15 per 1MGemma 4 31B IT: 58, $0.25 per 1MQwen3.6 Plus: 37, $1.12 per 1MGPT-5.4 nano: 48, $0.46 per 1MGPT-5.4 mini: 44, $1.69 per 1MGPT-5.4 Pro: 67, $67.5 per 1MGPT-5.4: 64, $5.62 per 1MGemini 3.1 Flash Lite: 49, $0.56 per 1MQwen3.5 Flash: 47, $0.18 per 1MQwen3.5 9B: 53, $0.19 per 1MQwen3.5 35B-A3B: 56, $0.69 per 1MGemini 3.1 Pro Preview: 55, $4.5 per 1MClaude Sonnet 4.6: 54, $6 per 1MQwen3.5 Plus: 63, $0.9 per 1MQwen3.5 397B-A17B: 52, $1.35 per 1MGLM-5: 60, $1.55 per 1MClaude Opus 4.6: 57, $10 per 1MGLM-4.7: 48, $1 per 1MGemini 3 Flash Preview: 50, $1.12 per 1MGPT-5.2 Pro: 61, $57.8 per 1MGPT-5.2 Codex: 51, $4.81 per 1MGPT-5.2: 54, $4.81 per 1MDeepSeek V3.2: 64, $0.32 per 1MClaude Opus 4.5: 51, $10 per 1MGPT-5.1 Codex Max: 37, $3.44 per 1MGPT-5.1: 64, $3.44 per 1MGPT OSS 120B: 64, $0.26 per 1MGPT OSS 120BPrice per 1M tokens →↑ Math

OpenRouter $0.15 / $0.6 · Cached input $0.075

The GPT OSS line

2 releases in the GPT OSS line.

GPT OSS 120BAug 5, 20250dGPT OSS 20BAug 5, 2025
ECI STANDING#123 of 270
SCORE139.79135.57–142.12 CI
FRONTIER LAGLaunched 7.6 months behindScores about what o1 did in December 2024.
SCORES LIKE THESE MODELS
Kimi K2 (Jul 2025)Moonshot · Jul 2025139.97
Gemini 2.5 Flash (Apr 2025)Google DeepMind · Apr 2025139.84
DeepSeek-V3.1DeepSeek · Aug 2025139.79
Qwen3-30B-A3B-Thinking (Jul 2025)Alibaba · Jul 2025139.55
Full Epoch Capabilities Index ranking ↗

Epoch benchmark boards

Best result on each board it has run. The tick marks the leader.

OTIS Mock AIME 2024-2025
89%#67 of 216
WeirdML
48%#56 of 141
METR Time Horizons
57%#21 of 42
DTBench
76%#96 of 179
Fiction.LiveBench
44%#48 of 58
Terminal Bench
19%#46 of 50
Surface Evolver Bench
25%#24 of 26
SimpleBench
22%#83 of 89
2 more boards
Mystery Game Puzzles
2%#88 of 88
APEX-Agents
4%#51 of 51

OpenAI's past year

24 release days in the past year, about one every 15 days. The next is expected around Nov 8, 2026. Forecasts ↗

NovDecJanFebMarAprMayJunJulAugSepOctNov

Specifications

RELEASE DATEAug 5, 2025
WEIGHTSOpen weights
CONTEXT WINDOW131.1K
MAX OUTPUT32.8K
REASONINGNot listed
TOOL CALLINGNot listed
INPUTStext
OUTPUTStext

Models.dev record ↗

Specifications and prices from Models.dev; availability and limits may differ by provider. Benchmarks from Epoch AI (CC BY 4.0), LiveBench and the other sources named above.