gpt-oss-20b

1318 Score third-party Elo · arena.ai / LMArena ?

gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...

LLM & Generative AI 1 model 5 listings snapshot Oct 14, 2026, 12:54 AM

Platforms

click a card to open the detail table

Cheapest OpenRouter no official benchmark Input$0.018 Output$0.090 per 1M tokens Fastest OpenRouter lowest p50 latency on this board 128.1 ms p50 Most free quota — largest published free allowance — free tier

Model variants

W3r9 dedup — same model + identical tags merge; prices are listed per billing form

Model	Provider	Tags	Input /1M	Output /1M
openai/gpt-oss-20b	OpenRouter	20b128k contextreasoning	$0.018	$0.090
openai/gpt-oss-20b:batch :batch	OpenRouter	20b128k contextreasoning	$0.024	$0.112
openai/gpt-oss-20b	DeepInfra	20b128k contextreasoning	$0.030	$0.140
openai/gpt-oss-20b	Together	20b128k context	$0.050	$0.200
openai/gpt-oss-20b	Groq	20b128k context	$0.075	$0.300

Capabilities

official design_arena & architecture metadata

Design Arena Elo official design_arena benchmark (OpenRouter)

dataviz 933 · 39.7% wr
website 855 · 27.9% wr

Input / Output modalities modality: text->text

text in+out

Platform details

per provider

provider	model	Input /1M	Output /1M	Latency ?	Official $/1M	Availability	Data time
DeepInfra	openai/gpt-oss-20b	$0.030	$0.140	—	—	—	2026-10-14 00:43:44
Groq	openai/gpt-oss-20b	$0.075	$0.300	—	—	—	2026-10-14 00:54:15
OpenRouter	openai/gpt-oss-20b:batch	$0.024	$0.112	128.1 ms	—	100% available	2026-10-14 00:43:21
Together	openai/gpt-oss-20b	$0.050	$0.200	—	—	—	2026-10-14 00:53:13

4 rows · snapshot 2026-10-14 00:54:15 · from data/public/prices.json + measurements.json