RivenGet started

Search docs

Find a page across the Riven documentation

Model fleet

Per-model reference generated from the live Riven gateway catalog. Every entry is callable through the API with the same key.

The gateway serves 75 live models — three on the on-prem fleet and 33 across nine cloud providers. The 65 entries below include the alias IDs that resolve onto those models, which is why the catalog is longer than the model count. See Riven Ultra Router for the mapping.

Riven (on-prem and routed aliases) 22

Model IDTier
glm-5.2frontier
nomic-embed-textembed
qwen3.6-35b-a3b-fp8frontier
riven-bestall plans
riven-codeall plans
riven-corecore
riven-embedall plans
riven-fastfast
riven-imageall plans
riven-instantinstant
riven-maxmax
riven-propro
riven-researchresearch
riven-visionvision
riven-agentall plans
riven-chat-lgall plans
riven-chat-mdstandard
riven-chat-smstandard
riven-chat-xlall plans
riven-embed-fastall plans
riven-reasonreason
rvn-assistant-v2custom

OpenAI 10

Model IDTier
gpt-4ocore
gpt-4o-minifast
gpt-5core
gpt-5.4-minifast
gpt-5.4-nanoinstant
gpt-5.5pro
gpt-5.6pro
gpt-5.6-lunafast
gpt-5.6-solpro
gpt-5.6-terracore

NVIDIA NIM 9

Mistral 7

Model IDTier
codestral-2508fast
mistral-large-2512core
mistral-large-latestcore
mistral-medium-3fast
mistral-medium-3.5core
mistral-medium-latestcore
mistral-small-latestinstant

Anthropic 5

Model IDTier
claude-fable-5max
claude-haiku-4-5fast
claude-opus-4-7pro
claude-opus-4-8pro
claude-sonnet-4-6core

Moonshot 5

Model IDTier
kimi-k2.5core
kimi-k2.6core
kimi-k2.7-codecore
kimi-k2.7-code-highspeedcore
kimi-k3pro

Perplexity 3

Model IDTier
sonarresearch
sonar-proresearch
sonar-reasoning-proresearch

Groq 2

Model IDTier
llama-3.1-8b-instantinstant
llama-3.3-70b-versatilefast

Fireworks 1

Model IDTier
gpt-oss-120bfast

Cerebras 1

Model IDTier
zai-glm-4.7instant