vlm.fyi Get the data

Reference

Models

Each card compares the providers serving one model: frames per request on a shared scale, and video support.

Qwen3.8 27B

Qwen · 22 providers

Alibaba Cloud via OpenRouter Video ≥ 512 MB 224
Jarvislabs Video up to 256 MB 4
AkashML (fp8) via OpenRouter Video: unknown —
Cerebras (fp16) via OpenRouter Video: unknown —
Chutes (fp8) via OpenRouter Video: unknown —
Cloudflare via OpenRouter Video: unknown —
CoreWeave (fp8) via OpenRouter Video: unknown —
DarkBloom (fp4) via OpenRouter Video: unknown —
DeepInfra (bf16) via OpenRouter Video: unknown —
DekaLLM via OpenRouter Video: unknown —
IonStream (fp8) via OpenRouter Video: unknown —
Mancer (fp8) via OpenRouter Video: unknown —
Modal via OpenRouter Video: unknown —
ModelRun (fp4) via OpenRouter Video: unknown —
NEAR AI (fp8) via OpenRouter Video: unknown —
Novita via OpenRouter Video: unknown —
Parasail (fp8) via OpenRouter Video: unknown —
Phala via OpenRouter Video: unknown —
Reka (fp8) via OpenRouter Video: unknown —
Tenstorrent (fp4) via OpenRouter Video: unknown —
Venice (fp8) via OpenRouter Video: unknown —
Wafer (fp8) via OpenRouter Video: unknown —

Qwen3.6 35B A3B

Qwen · 11 providers

Auto-routed via OpenRouter Video ≥ 512 MB 160
AkashML (fp8) via OpenRouter Video: unknown —
Atlas Cloud (fp8) via OpenRouter Video: unknown —
CoreWeave (fp8) via OpenRouter Video: unknown —
DarkBloom (fp4) via OpenRouter Video: unknown —
DeepInfra (fp8) via OpenRouter Video: unknown —
DekaLLM via OpenRouter Video: unknown —
Parasail (fp8) via OpenRouter Video: unknown —
Phala via OpenRouter Video: unknown —
SiliconFlow (fp8) via OpenRouter Video: unknown —
Venice (fp8) via OpenRouter Video: unknown —

Gemini 3.7 Flash

Google · 7 providers

Google AI Studio via OpenRouter Video up to 12 MB ≥ 384
Google AI Studio (flex) via OpenRouter Video: unknown —
Google AI Studio (priority) via OpenRouter Video: unknown —
Google Vertex (global) via OpenRouter Video: unknown —
Google Vertex (global, flex) via OpenRouter Video: unknown —
Google Vertex (global, priority) via OpenRouter Video: unknown —
Google Vertex (us) via OpenRouter Video: unknown —

DeepSeek V4.1 Flash (1017)

DeepSeek · 5 providers

Jarvislabs No video ≥ 384
Modal No video ≥ 384
Together AI No video ≥ 384
Fireworks via OpenRouter No video 112
Fireworks No video 112

GLM 5.3 Flash

Z.ai · 4 providers

Together AI Video up to 256 MB ≥ 384
Z.ai via OpenRouter No video ≥ 384
Jarvislabs Video up to 64 MB 128
Fireworks No video 112

Kimi K3

MoonshotAI · 3 providers

Modal No video ≥ 384
Together AI Video ≥ 512 MB ≥ 384
Fireworks Video up to 96 MB 112

Gemma 4 31B

Google · 1 provider

Jarvislabs Video up to 256 MB 4

Muse Spark (open)

Meta · 1 provider

Fireworks No video 112