gpt-oss-120b vs gpt-oss-20b
At a Glance
| Compare |  gpt-oss-120bOpenAI | gpt-oss-20bOpenAI |
|---|---|---|
| Intelligence, Cost, and Efficiency | ||
| IntelligenceHigher is better · MM Intelligence v4.4 | #43 of 570.0 score · 1/3 sources · single-source provisional · missing LiveBench, ARC-AGI-2 · full-core range 0.0–66.7 | #43 of 570.0 score · 1/3 sources · single-source provisional · missing LiveBench, ARC-AGI-2 · full-core range 0.0–66.7 |
| Pricing and Limits | ||
| Input priceFrom · USD / 1M tokens | $0.0296Openrouter ↗ · Oct 19, 2026 | $0.018Openrouter ↗ · Oct 19, 2026 |
| Output priceFrom · USD / 1M tokens | $0.136Openrouter ↗ · Oct 19, 2026 | $0.090Openrouter ↗ · Oct 19, 2026 |
| Context windowMaximum documented tokens | 131K | 131K |
| Model facts checked | Aug 28, 2026View model evidence → | Aug 28, 2026View model evidence → |
Token prices are the lowest available sourced USD rates; input and output may use different providers. Cost ranking estimates output spend on LiveBench, not a full request bill. Ranking methodology →
Available Benchmarks
| Benchmark | gpt-oss-120b | gpt-oss-20b |
|---|---|---|
| LMArena Text Arenatext-2026-10-15-a64ca3b12b6e · arena_rating · leader | 1,365.51100% of row best · rating · gpt-oss-120b; 95% CI [1361.16495776, 1369.85717870]; votes 30206; rank 192 | 1,287.4394% of row best · rating · gpt-oss-20b; 95% CI [1281.03637900, 1293.81371199]; votes 10419; rank 260 |
| Overall ResultCounted from the protocol-matched rows above | 1 benchmark winNo overall winner | 0 benchmark winsNo overall winner |
Third-party benchmark Only like-for-like primary-publisher results are shown; raw scores, relative scores, configuration, and token spend remain visible.
Side-by-Side Facts
| Field | gpt-oss-120b | gpt-oss-20b |
|---|---|---|
| Developer | OpenAI | OpenAI |
| Family | Gpt Oss 120b | Gpt Oss |
| Model | gpt-oss-120b | gpt-oss-20b |
| Version | gpt-oss-120b | gpt-oss-20b |
| Lifecycle | active | active |
| Released | Unknown | Unknown |
| Knowledge cutoff | Unknown | Unknown |
| Input modalities | Text | Text |
| Output modalities | Text | Text |
| Context window | 131K | 131K |
| Total parameters | 116.8B | 20.9B |
| Active parameters | 5.1B | 3.6B |
| License | apache-2.0 | apache-2.0 |
| Open weights | Yes | Yes |
| API available | Yes | Yes |
| Self-hostable | Yes | Yes |
| Provider access | Cerebras (Standard), Deepinfra (Standard), Fireworks Ai (Serverless, Standard), Groq (Standard), Hugging Face (Standard), Openrouter (Standard), Together Ai (Serverless, Standard) | Deepinfra (Standard), Groq (Standard), Hugging Face (Standard), Openrouter (Standard), Together Ai (Serverless, Standard) |
| Capabilities | chat, generation, reasoning, tools | chat, generation, reasoning, tools |
gpt-oss-120b Capabilities
gpt-oss-20b Capabilities
Related Comparisons
All text comparisons →| A | Pair | B | Context |
|---|---|---|---|
| vs | gpt-oss-120bOpenAI | cross-developer peerstext | |
 Phi-4-reasoningMicrosoft | vs | gpt-oss-120bOpenAI | cross-developer peerstext |
 Llama-3.1-70BMeta | vs | gpt-oss-120bOpenAI | cross-developer peerstext |
| vs | gpt-oss-120bOpenAI | cross-developer peerstext | |
| vs | gpt-oss-120bOpenAI | cross-developer peerstext | |
 Devstral-2-123B-Instruct-2512Mistral AI | vs | gpt-oss-120bOpenAI | cross-developer peerstext |
| vs | gpt-oss-120bOpenAI | cross-developer peerstext | |
| vs | gpt-oss-120bOpenAI | cross-developer peerstext | |
GPT-4.1OpenAI | vs | gpt-oss-120bOpenAI | family variantstext |
GPT-4oOpenAI | vs | gpt-oss-120bOpenAI | family variantstext |
GPT-5.6 SolOpenAI | vs | gpt-oss-120bOpenAI | family variantstext |
Primary Evidence
Sources and Freshness
Questions
gpt-oss-120b vs gpt-oss-20b FAQs
Is gpt-oss-120b or gpt-oss-20b better for coding?+
This comparison does not currently contain a protocol-matched coding benchmark for both gpt-oss-120b and gpt-oss-20b, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.
Which is cheaper, gpt-oss-120b or gpt-oss-20b?+
gpt-oss-120b is $0.0296 and gpt-oss-20b is $0.018 per million tokens, so gpt-oss-20b is cheaper on this metric. gpt-oss-120b is $0.136 and gpt-oss-20b is $0.090 per million tokens, so gpt-oss-20b is cheaper on this metric.
Which has a larger context window, gpt-oss-120b or gpt-oss-20b?+
Both models have the same sourced context window: 131,072 tokens each.
Which performs better in benchmarks, gpt-oss-120b or gpt-oss-20b?+
There is no overall benchmark winner: At least two independently verified, protocol-matched benchmarks are required for an overall winner.
Can gpt-oss-120b or gpt-oss-20b be self-hosted?+
Both models have the same sourced self-hosting status: supported. gpt-oss-120b is open weight; gpt-oss-20b is open weight.
Can gpt-oss-120b and gpt-oss-20b understand images?+
gpt-oss-120b is not documented with image input; gpt-oss-20b is not documented with image input. This reflects supported input modalities, not vision quality.
Which can generate longer answers, gpt-oss-120b or gpt-oss-20b?+
Neither model has a sourced maximum output in this comparison, so a larger limit cannot be determined.
Do gpt-oss-120b and gpt-oss-20b support reasoning and tool use?+
gpt-oss-120b: reasoning and tool calling. gpt-oss-20b: reasoning and tool calling. Feature support does not establish relative quality.
Which is available from more inference providers, gpt-oss-120b or gpt-oss-20b?+
gpt-oss-120b has 7 sourced provider routes; gpt-oss-20b has 5, so gpt-oss-120b has broader tracked availability.
Which offers better value, gpt-oss-120b or gpt-oss-20b?+
There is no universal value winner. Compare the input and output prices above with the matched benchmark result for your workload: cheaper tokens can be offset by different quality, token usage, latency, or provider availability.