sarvamai / text-generation updated 5 months ago

sarvam-30b

1. Introduction 2. Architecture 3. Benchmarks - Knowledge & Coding - Reasoning & Math - Agentic 4. Inference - Hugging Face - vLLM - SGLang 5. Footnote 6. Citation

Params
32.2 B
Context
131,072
Downloads 30d
304 K
Likes
225
Commercial use: allowed apache-2.0 Not gated SAFETENSORS 23 languages View on Hugging Face ↗

Download history

daily snapshots · 11 days
304 K237 K
Sep 11Sep 14Sep 18Sep 21

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

FileQuantSizeEst. VRAMVerdict on RTX 4090 · 24 GB
model.safetensors f32 128.6 GB 146.8 GB ❌ Won’t fit
model.safetensors (bf16, full) bf16 + 128K ctx 128.6 GB 219.1 GB ❌ Won’t fit
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Run it

copy-paste, exact tags checked against the Hub
~ · curl · api/v1
$ curl -s https://aimodelscomparison.com/api/v1/models/sarvam-30b
{
  "hf_id": "sarvamai/sarvam-30b",
  "params_b": 32.15,
  "context_length": 131072,
  "license": { "id": "apache-2.0", "commercial": "yes" },
  "downloads_30d": 304336,
  "vram_estimates": [
    { "quant": "f32", "gb": 146.8 }
  ],
  "updated_at": "2026-09-11T01:00:33Z"
}
est. VRAM —on RTX 4090 · 24 GBJSON API →

Specifications

Architecture
SarvamMoEForCausalLM
Parameters
32.2 B
Tensor type
F32
Context length
131,072
Vocabulary
262,144
Layers / heads
19 / 64
Licence
apache-2.0
First seen on the Hub
2026-03-03
Training datasets
undisclosed
Added to our catalog
2026-09-11
Compare with any text-generation model