vram.run Models Hardware Providers Cloud State of Inference

API provider data is live · Hardware & cloud pricing curated 2026-02-23

Featherless vs Novita

207 vs 67 models, 44 shared

Shared models

Model	Featherless $/1M out	Featherless tok/s	Novita $/1M out	Novita tok/s
MiniMax-M2				86 tok/s
MiniMax-M2.1				72 tok/s
MiniMax-M2.5				67 tok/s
MiniMax-M2.7				40 tok/s
MiniMax-M3				39 tok/s
Qwen2.5-72B-Instruct				27 tok/s
Qwen3-235B-A22B				16 tok/s
Qwen3-235B-A22B-Thinking-2507				65 tok/s
Qwen3-Coder-480B-A35B-Instruct				48 tok/s
Qwen3-Coder-Next				120 tok/s
Qwen3-Next-80B-A3B-Instruct				89 tok/s
Qwen3-VL-235B-A22B-Thinking				43 tok/s
Qwen3-VL-30B-A3B-Instruct				102 tok/s
Qwen3-VL-8B-Instruct				73 tok/s
Qwen3.5-27B				68 tok/s
Qwen3.5-397B-A17B				73 tok/s
L3-8B-Lunaris-v1				71 tok/s
L3-8B-Stheno-v3.2				81 tok/s
DeepSeek-R1-0528				25 tok/s
DeepSeek-R1-Distill-Llama-70B				60 tok/s
DeepSeek-V3-0324				37 tok/s
DeepSeek-V3.1				25 tok/s
DeepSeek-V3.1-Terminus				27 tok/s
DeepSeek-V3.2				39 tok/s
DeepSeek-V4-Flash				86 tok/s
DeepSeek-V4-Pro				59 tok/s
gemma-4-26B-A4B-it				34 tok/s
gemma-4-31B-it				20 tok/s
Llama-3.1-8B-Instruct				163 tok/s
Llama-3.3-70B-Instruct				34 tok/s
Meta-Llama-3-70B-Instruct				23 tok/s
Kimi-K2-Instruct				25 tok/s
Kimi-K2-Instruct-0905				25 tok/s
Kimi-K2-Thinking				62 tok/s
Kimi-K2.5				57 tok/s
Kimi-K2.6				84 tok/s
Kimi-K2.7-Code				47 tok/s
gpt-oss-120b				134 tok/s
gpt-oss-20b				101 tok/s
GLM-4.6				57 tok/s
GLM-4.7				57 tok/s
GLM-4.7-Flash				57 tok/s
GLM-5				35 tok/s
GLM-5.2				47 tok/s

Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0