bansos-router models

38 free coding models across 4 upstreams, as the daemon's own catalog check saw them.

Checked from GitHub Actions, every six hours. Some upstreams answer differently by region and IP, so run bansos models for what your machine can reach.

OpenCode Zen 8

model idreasoningvisioncontextmax output
nemotron-3.5-lightning-freeyesno262k262k
beam-freeyesno1M131k
flint-40b-freeyesno262k66k
nemotron-3.5-super-vl-freeyesyes1M131k
big-pickleyesno200k32k
deepseek-v4.1-flash-freeyesyes200k32k
glm-5.3-flash-freeyesyes200k32k
mimo-v2.6-flash-freeyesyes200k32k

KiloCode gateway 15

model idreasoningvisioncontextmax output
kilo-auto/freeyesno256k10k
nvidia/nemotron-3.5-super-vl:freeyesyes1M10k
recursive/flint-v1.0:freeyesno262k10k
xingchen-agi/xing4.0-29b-a4b:freeyesno262k10k
dots-studio/dots-3-note-preview:freeyesyes512k66k
liquid/lfm-2.5-2.6b:freeyesno66k8k
nvidia/nemotron-3.5-lightning:freeyesno1M66k
poolside/laguna-s-2.1:freeyesno262k33k
poolside/laguna-xs-2.1:freeyesno262k33k
cohere/north-mini-code:freeyesno256k64k
nvidia/nemotron-3.5-content-safety:freeyesyes128k8k
nvidia/nemotron-3-ultra-550b-a55b:freeyesno1M66k
nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:freeyesyes256k66k
nvidia/nemotron-3-super-120b-a12b:freeyesno262k262k
openrouter/freeyesyes200k66k

LLM7 7

model idreasoningvisioncontextmax output
codestral-latestnono32k8k
gpt-oss:20bnono128k16k
minimax-m2.7yesno180k33k
mistral-Nemo-Instruct-2407nono128k16k
nemotron-3-nano:30byesno1M16k
defaultnono128k8k
fastnono128k8k

OVHcloud AI Endpoints 8

model idreasoningvisioncontextmax output
ovh/gpt-oss-120byesno131k33k
ovh/gpt-oss-20byesno131k33k
ovh/Qwen3.5-397B-A17Byesyes262k66k
ovh/Qwen3.8-27Byesyes262k33k
ovh/Qwen3.6-27Byesyes262k66k
ovh/Qwen3.5-9Byesyes262k66k
ovh/Mistral-Small-3.2-24B-Instruct-2506noyes131k16k
ovh/Meta-Llama-3_3-70B-Instructnono131k16k