BF16 GGUF models

Unquantized bfloat16. Original quality and the largest size. 9004 models have a BF16 file. Sizes run from 1M to 2780B parameters. The longest trained context is 10240K tokens.

Most downloadedNewestLargestLongest contextSmallest
QW
Qwen3 4B Thinking 2507 GPT 5.2 High Reasoning Distill
Qwen · TeichAI
Params4.02B
Context256K
Q4_K_M2.5 GB
ReasoningTool use
GR
granite-3b-code-base-128k
Granite · Mungert
Params3.48B
Context125K
Q4_K_M2.1 GB
Coding
LI
LightNav 0
Lightnav · prithivMLmods
Params4.43B
Context256K
Q4_K_M2.7 GB
Tool useVision
LL
KernelLLM
Llama · Mungert
Params8.03B
Context128K
Q4_K_M5.0 GB
GGUF
QW
Qwen3.6 35B A3B
Qwen · ggml-org
Params447M
Context256K
Smallest37.8 GB
Vision
GE
Gemma-3-12B-It
Gemma · Abc7347
Params11.8B
Context128K
Q4_K_M7.3 GB
Vision
QW
Qwen2.5 VL 7B NSFW Caption V3
Qwen · Hi821119
Params7.62B
Context125K
Q4_K_M4.7 GB
Vision
QW
DMind-1-mini
Qwen · Mungert
Params14.8B
Context40K
Q4_K_M9.1 GB
GGUF
BO
Borealis Open 1b
Borealis · NbAiLab
Params1B
Context32K
Q4_K_M0.8 GB
GGUF
PO
Polaris 4B Preview
Polaris · bartowski
Params4.02B
Context128K
Q4_K_M2.5 GB
GGUF
DA
DASD-4B-Thinking
Dasd · Mungert
Params4.02B
Context256K
Q4_K_M2.5 GB
ReasoningTool use
NE
nvidia OpenCodeReasoning Nemotron 7B
Nemotron · bartowski
Params7.62B
Context32K
Q4_K_M4.7 GB
CodingReasoningTool use
MA
Mario Html v2 9B
Mario · prithivMLmods
Params9.2B
Context256K
Q4_K_M5.8 GB
ReasoningTool useVision
TI
tiny1m
Tiny · shibatch
Params—
Context256
Q4_K_M0.0 GB
GGUF
DE
DeepSeek-R1-Distill-Qwen-14B
DeepSeek · Mungert
Params14.8B
Context128K
Q4_K_M9.0 GB
ReasoningTool use
CO
CohereLabs.command A Plus 05 2026 Bf16
Command · DevQuasar
Params218B
Context4883K
Smallest79.7 GB
Tool use
CO
context 1
Context · nicolasembleton
Params20.9B
Context128K
Q4_K_M15.8 GB
Tool use
NE
Next 1b
Next · thelamapi
Params1B
Context32K
Smallest1.1 GB
Coding
QO
Qolda
Qolda · issai
Params4.02B
Context40K
Q4_K_M2.5 GB
ReasoningTool useVision
QW
Qwen3.5 0.8B
Qwen · prithivMLmods
Params773M
Context256K
Q4_K_M0.5 GB
ReasoningTool useVision
LO
Logics-Parsing
Logics · Mungert
Params7.62B
Context125K
Q4_K_M5.0 GB
GGUF
GR
Granite 4.0 H Small Heretic High
Granite · rodrigomt
Params32.2B
Context1024K
Smallest18.8 GB
Tool use
GE
Gemma-3-4B-It
Gemma · Abc7347
Params3.88B
Context128K
Q4_K_M2.5 GB
Vision
NO
Normistral 11b Thinking 4
Normistral · ltg
Params11.4B
Context1000K
Q4_K_M6.9 GB
Reasoning
CA
Capek0.5 35B A3B
Capek · prithivMLmods
Params34.7B
Context256K
Smallest19.8 GB
ReasoningTool useVision
DE
Nemotron Orchestrator 8B DeepSeek v3.2 Speciale Distill
DeepSeek · TeichAI
Params8.19B
Context40K
Q4_K_M5.0 GB
ReasoningTool use
WE
WeVisDoc 4B
Wevisdoc · prithivMLmods
Params4.41B
Context256K
Q4_K_M2.7 GB
Tool useVision
LL
Llama-3.2-3B-Instruct
Llama · mny1238
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
ZW
ZwZ2 27B
Zwz · prithivMLmods
Params26.9B
Context256K
Q4_K_M16.5 GB
ReasoningTool useVision
KA
KAT Coder V2.5 Dev 35B A3B MTP ABLITERATED
Kat · jakeroxs
Params35.5B
Context256K
Q4_K_M21.7 GB
CodingReasoningTool use

Other quants

Keep browsing