Sunk Cost sunkcost.ai Data checked 2026-09-03

Qwen3-Coder Next vs gpt-oss-20b

Qwen3-Coder Next scores higher on the intelligence index — 10 against 9. gpt-oss-20b is the smaller download at 12 GB, so it runs on cheaper hardware.

Qwen3-Coder Nextgpt-oss-20b
Intelligence index109
ClassBelow every hosted tierBelow every hosted tier
Weights48 GB12 GB
QuantisationQ4_K_MMXFP4
Parameters79.7B (3B active)20.9B (3.6B active)
Max context256k128k
API price per 1M$0.12 in / $0.8 out$0.02 in / $0.1 out
LicenceApache 2.0Apache 2.0
Cheapest machine that runs itStrix Halo Framework Desktop, 128GB $3,449Mac mini M6, 24GB $1,099
Summarising not rated good
Translation not rated usable
Everyday coding not rated good
Reasoning & maths not rated usable
Agentic work not rated usable

Ratings are coarse on purpose. Speeds and pay-back depend on the machine — open either model's page for the full list, or see both against the frontier. Context is 32k throughout.