Sarvam GGUF models

111 GGUF models in the Sarvam family, quantized by 56 uploaders. Sizes run from 13M to 106B parameters. The longest trained context is 256K tokens.

TinySmallMediumLarge
Most downloadedNewestLargestLongest contextSmallest
SA
Sarvam Translate
Sarvam · Savyasaachin
Params3.88B
Context128K
Smallest3.2 GB
GGUF
SA
Sarvam Translate
Sarvam · SrijitSet99
Params3.88B
Context128K
Q4_K_M2.5 GB
GGUF
SA
Sarvam 2b v0.5
Sarvam · rachittshah
Params2.51B
Context8K
Q4_K_M1.5 GB
GGUF
SA
Sarvam M
Sarvam · tifin-india
Params23.6B
Context32K
Smallest8.9 GB
Reasoning
SA
Sarvam 1
Sarvam · singarajusaiteja
Params2.53B
Context8K
Smallest5.1 GB
Tool use
SA
Sarvam M
Sarvam · Savyasaachin
Params23.6B
Context32K
Smallest11.5 GB
Reasoning
SA
Sarvam 2b v0.5
Sarvam · Ffftdtd5dtft
Params2.51B
Context8K
Smallest1.0 GB
GGUF
SA
guhan m q8
Sarvam · mygitphase
Params23.6B
Context32K
Smallest25.1 GB
Reasoning
SA
Sarvam 1
Sarvam · Nithin-2812
Params2.53B
Context8K
Q4_K_M1.5 GB
Tool use
SA
sarvam model
Sarvam · mayupat13
Params2.53B
Context8K
Q4_K_M1.5 GB
Tool use
SA
Sarvam Translate
Sarvam · mashriram
Params3.88B
Context128K
Smallest2.5 GB
GGUF
SA
sarvam_instruct
Sarvam · Savyasaachin
Params13M
Context—
Smallest0.1 GB
GGUF
SA
sarvam_instruct
Sarvam · Savyasaachin
Params13M
Context—
Smallest0.1 GB
GGUF
SA
Sarvam Translate
Sarvam · mashriram
Params3.88B
Context128K
Smallest7.8 GB
GGUF
SA
merged_sarvamInstructHindi
Sarvam · aman313
Params2.53B
Context8K
Smallest5.1 GB
Tool use
SA
Sarvam 30b
Sarvam · kmr-kishore
Params32.2B
Context128K
Q4_K_M19.5 GB
ReasoningTool use
SA
Sarvam Translate
Sarvam · mashriram
Params3.88B
Context128K
Smallest7.8 GB
GGUF
SA
Sarvam 2b v0.5
Sarvam · rumbleFTW
Params2.51B
Context8K
Smallest2.7 GB
GGUF
SA
Sarvam 1
Sarvam · DevBM
Params2.53B
Context8K
Smallest5.1 GB
Tool use
SA
guhan m
Sarvam · mygitphase
Params23.6B
Context32K
Smallest47.2 GB
Reasoning
SA
Sarvam 2B GGUF llm
Sarvam · Vikasith
Params2.51B
Context8K
Smallest5.0 GB
GGUF
SA
Sarvam 30b
Sarvam · kmr-kishore
Params32.2B
Context128K
Smallest64.3 GB
ReasoningTool use
SA
Sarvam 30B
Sarvam · WizUser
Params32.2B
Context128K
Smallest20.7 GB
ReasoningTool use
SA
sarvam 30b compressed
Sarvam · kmr-kishore
Params32.2B
Context128K
Q4_K_M19.5 GB
ReasoningTool use

Keep browsing