← Model list

GPT OSS 20B

openai·openai/gpt-oss-20b·Deprecated·Open weights·gpt-oss series

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

At a glance
Reference price$0.40 / $1.80 per 1M
Cache read · Regolo AI
Lowest paid$0.03 / $0.13
OpenRouter Gateway · 13.3× spread
4 more $0 channels
Context128,000
⚠ Providers report 64,000–131,072; the table below is authoritative
Output limit16,384
Capabilities
ReasoningTool useStructured outputTemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
Text
Knowledge cutoff2024-10
Released / updated2025-08-05 / 2025-08-05

Quality & performanceArtificial Analysis · Intelligence Index v4.1 · rep. tier high

Intelligence15.2
Value276
Coding20.7
Agentic3.1
Output speed136 tok/s
TTFT0.98 s
Cost per task$0.0250
Value formulaIQ 15.2 ÷ min blended $0.055 = 276
Reasoning tier → intelligence / speed (higher tier = stronger but slower)
lowIQ 14.4 · 170 tok/s
highIQ 15.2 · 136 tok/s

Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.

Available at 45 providers39 with public prices · 4 free

ProviderTierInputOutputCache readCache writeContextOutput limitStatus
NvidiaFirst-partyFree131,07232,768
QVACGatewayFree131,07232,768
KenariGatewayFree131,07232,768
LMStudioCloudFree131,07232,768
OpenRouterGateway$0.03$0.13$0.03131,072131,072
Eden AIGateway$0.03$0.13131,07232,768
Kilo GatewayGateway$0.03$0.13$0.03131,072131,072
Weights & BiasesCloud$0.03$0.13$0.03131,072131,072
Deep InfraCloud$0.03$0.14131,07216,384
IO.NETCloud$0.03$0.14$0.015$0.0664,0004,096

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Your usage cost

1IO.NET$11.20
2OpenRouter$12.50
3Eden AI$12.50
4Kilo Gateway$12.50
5Weights & Biases$12.50
6Deep Infra$13.00

4 more channels offer $0 (Nvidia, QVAC, Kenari etc.); free tiers usually have rate limits and no SLA, excluded from ranking.

The cheapest paid channel is the only channel.

Benchmark

No upstream benchmark data for this model. For quality, see the Artificial Analysis intelligence score above.

Reasoning control

effort = loweffort = mediumeffort = high

7 / 45 providers expose no reasoning control (reasoning_options: []).

Related models

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$1.80$02026-08-052026-08-132026-08-05 · Input list $0.402026-08-06 · Input list $0.402026-08-07 · Input list $0.402026-08-08 · Input list $0.402026-08-09 · Input list $0.402026-08-10 · Input list $0.402026-08-11 · Input list $0.402026-08-12 · Input list $0.402026-08-13 · Input list $0.402026-08-05 · Output list $1.802026-08-06 · Output list $1.802026-08-07 · Output list $1.802026-08-08 · Output list $1.802026-08-09 · Output list $1.802026-08-10 · Output list $1.802026-08-11 · Output list $1.802026-08-12 · Output list $1.802026-08-13 · Output list $1.802026-08-05 · Min blended $0.0552026-08-06 · Min blended $0.0552026-08-07 · Min blended $0.0552026-08-08 · Min blended $0.0552026-08-09 · Min blended $0.0552026-08-10 · Min blended $0.0552026-08-11 · Min blended $0.0552026-08-12 · Min blended $0.0552026-08-13 · Min blended $0.055