sarvam-105b
sarvam-105b is an open-weight language model from Sarvam AI with 106B parameters (mixture of experts). At 4-bit it needs about 112.5 GB of memory, so the smallest current device that fits it is the H200 SXM.
Overview
Includes room for an 8k context and runtime overhead. More context needs more memory.
Key information
- Lab
- Sarvam AI
- Released
- 3 Mar 2026
- Parameters
- 106B · MoE
- Type
- Language
- Context
- 131,072 tokens
- License
- Apache-2.0
- HF downloads, 30 days
- 5,176
- Repository
- sarvamai/sarvam-105b
More from Sarvam AI
View allWill it run on your hardware?
How we estimate| Estimate or device | Memory |
|---|---|
| Q4 | 112.5 GB |
| Q8 | 166.2 GB |
| BF16 | 275.8 GB |
| RTX 5060 Ti 16 GB | 16 GB usable |
| RTX 5090 32 GB | 32 GB usable |
| DGX Spark 128 GB | 108.8 GB usable |
| Mac Studio M5 Ultra 256 GB | 192 GB usable |
Single-device memory estimate. Pick a device to update the On your device column.
36 GB usable memory
| Format | Memory needed | On your device | Example hardware |
|---|---|---|---|
| 4-bit (Q4) | ≈ 112.5 GB | Too big 313% | H200 SXM141 GB memory |
| 8-bit (Q8) | ≈ 166.2 GB | Too big 462% | Mac Studio M5 Ultra (256GB)256 GB memory |
| 16-bit (BF16) | ≈ 275.8 GB | Too big 766% | Mac Studio M3 Ultra (512GB)Discontinued512 GB memory |
Fits means the estimate stays within 90% of usable memory. Tight is above 90% and up to 100%. Too big is above 100%. These are memory checks, not speed or compatibility tests.
Mac allowances are 66% of total RAM up to 36 GB, and 75% above. AI boxes allow 85% because their memory is shared with the operating system.
Where to run it
Through an API
No hosted API found.
More about this model
Frequently asked questions
How much memory does sarvam-105b need? At an 8k context, about 112.5 GB at 4-bit, 166.2 GB at 8-bit and 275.8 GB at 16-bit (BF16). A longer context needs more.
Can I run sarvam-105b on a Mac? The smallest current Mac that fits the 4-bit estimate of 112.5 GB is the Mac Studio M5 Ultra (256GB). We have not found a trusted MLX or GGUF build yet.
Does the full context fit? sarvam-105b supports up to 131,072 tokens of context. Our memory estimates use 8k, so long prompts need more memory than shown.
What does it cost through an API? We did not find a hosted API for this model.
License in plain English
Apache-2.0. This license permits commercial use, modification and redistribution, with conditions such as keeping the required notices. This summary is not legal advice.
OMHQ RigBench
Not yet tested. We show no speed or quality score without a measurement.
Revision history and sources
Last updated . Sources:Hugging Face model card.