AxForge
Sign in Get started

Model catalogue

Explore models

Model families from the labs, their builds and hardware requirements.

Model type

Filters apply together. Press Enter in search or choose Apply filters.

Reset filters GPU catalogue · AxForge model offerings

1095 matching of 3295 lab model families · Showing 751 to 800

Page 16 of 22

Model reference catalogue
Model TaskSizeContext LicenceEst. GPU Released
F2LLM codefuse-ai · 8 builds F2LLM-v2-14B Embeddings14.0B 40,960 tokens Apache 2.0 ~29.4 GiB2026-03-10
granite-speech-4.1-nar IBM · 1 build granite-speech-4.1-2b-nar Embeddings2.3B Apache 2.0 ~5.7 GiB2026-03-10
NVIDIA-Nemotron-3-Super NVIDIA · 3 builds NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 Text gen · MoE67.2B Licence fee ~86.7 GiB2026-03-10
PP-Chart2Table_safetensors Baidu · 1 build PP-Chart2Table_safetensors Image→text561M 32,768 tokens Apache 2.0 ~3.9 GiB2026-03-10
FLUX.2-klein-kv Black Forest Labs · 2 builds FLUX.2-klein-9b-kv-fp8 Image edit Licence fee ~15.5 GiB2026-03-09
s2-pro fishaudio · 1 build s2-pro Text→speech4.6B Licence fee ~12.8 GiB2026-03-09
Penguin-Encoder tencent · 1 build Penguin-Encoder Embeddings441M 40,960 tokens Apache 2.0 ~2.0 GiB2026-03-05
gemma-3-qat-q4_0-unquantized Lightricks · 1 build gemma-3-12b-it-qat-q4_0-unquantized Vision + text12.2B Conditions ~29.1 GiB2026-03-04
LTX-2.3 Lightricks · 4 builds LTX-2.3 Image→video Licence fee ~47.3 GiB2026-03-04
Cosmos-H-Surgical NVIDIA · 1 build Cosmos-H-Surgical Image→video15.2B Licence fee ~32.5 GiB2026-03-03
CubeComposer TencentARC · 1 build CubeComposer Video edit ~10.7 GiB2026-03-03
dasheng-denoiser mispeech · 1 build dasheng-denoiser Audio→audio119M Apache 2.0 ~1.9 GiB2026-03-03
granite-4.0-vision IBM · 1 build granite-4.0-3b-vision Vision + text4.0B 131,072 tokens Apache 2.0 ~12.2 GiB2026-03-03
sarvam sarvamai · 5 builds sarvam-30b Text gen32.2B 131,072 tokens Apache 2.0 ~66.7 GiB2026-03-03
FireRed-Image-Edit-1.1 FireRedTeam · 1 build FireRed-Image-Edit-1.1 Image edit20.4B Apache 2.0 ~47.0 GiB2026-03-02
LocateAnything NVIDIA · 1 build LocateAnything-3B Vision + text3.8B Licence fee ~10.1 GiB2026-03-02
Step-3.5-Flash-Midtrain stepfun-ai · 1 build Step-3.5-Flash-Base-Midtrain Text gen197.8B 131,072 tokens Apache 2.0 ~410.8 GiB2026-03-02
XCurOS0.1 XCurOS · 1 build XCurOS0.1-8B-Instruct Text gen7.6B 32,768 tokens Licence fee ~16.7 GiB2026-02-28
granite-4.0-speech IBM · 1 build granite-4.0-1b-speech Speech→text2.3B Apache 2.0 ~6.3 GiB2026-02-27
Infinity-Parser2-Flash infly · 1 build Infinity-Parser2-Flash Vision + text2.2B 262,144 tokens Apache 2.0 ~6.7 GiB2026-02-27
MediX-R1 MBZUAI · 4 builds MediX-R1-30B-GGUF Vision + text Licence fee 2026-02-27
Qwen3.5 Qwen · 30 builds Qwen3.5-9B Vision + text · MoE9.7B Apache 2.0 ~22.1 GiB2026-02-27
sd-amdnpu Stability AI · 1 build sd-turbo-amdnpu Text→image Licence fee ~7.0 GiB2026-02-25
sdxl-amdnpu Stability AI · 1 build sdxl-base-amdnpu Text→image Open ~8.4 GiB2026-02-25
Spatial-SSRL internlm · 2 builds Spatial-SSRL-3B Vision + text4.1B Apache 2.0 2026-02-25
XCurOS1.2-VLBF16 XCurOS · 1 build XCurOS1.2-8B-VLBF16-Instruct Vision + text8.8B 262,144 tokens Licence fee ~21.0 GiB2026-02-25
sooktam2 bharatgenai · 1 build sooktam2 Text→speech ~3.0 GiB2026-02-24
splade-code naver · 1 build splade-code-06B Embeddings596M 40,960 tokens Licence fee ~2.3 GiB2026-02-24
MedMO-Next MBZUAI · 2 builds MedMO-8B-Next Vision + text8.8B Apache 2.0 ~21.0 GiB2026-02-23
BSCodec espnet · 1 build BSCodec Audio→audio Apache 2.0 ~3.3 GiB2026-02-22
Falcon-OCR tiiuae · 1 build Falcon-OCR Image→text270M 8,192 tokens Apache 2.0 2026-02-22
Aurora-Spec-Minimax-M2.5 togethercomputer · 1 build Aurora-Spec-Minimax-M2.5 Text gen858M 196,608 tokens Apache 2.0 ~2.5 GiB2026-02-19
ColBERT-Zero lightonai · 1 build ColBERT-Zero Embeddings149M 8,192 tokens Apache 2.0 ~1.4 GiB2026-02-19
Cosmos-H-Surgical-Simulator NVIDIA · 1 build Cosmos-H-Surgical-Simulator Image→video Licence fee ~4.9 GiB2026-02-19
Olmo-Hybrid-SFT allenai · 1 build Olmo-Hybrid-Instruct-SFT-7B Text gen7.4B 32,768 tokens Apache 2.0 ~16.8 GiB2026-02-19
OpenEarthAgent MBZUAI · 1 build OpenEarthAgent Text gen4.0B 262,144 tokens Apache 2.0 ~10.1 GiB2026-02-19
qwen3-asr-f16 OpenVoiceOS · 1 build qwen3-asr-0.6b-f16 Speech→text ~3.5 GiB2026-02-19
qwen3-asr-q8-0 OpenVoiceOS · 1 build qwen3-asr-0.6b-q8-0 Speech→text ~3.0 GiB2026-02-19
qwen3-forced-aligner-q4-k-m OpenVoiceOS · 1 build qwen3-forced-aligner-0.6b-q4-k-m Speech→text ~2.3 GiB2026-02-19
glm4-moe-tiny PrimeIntellect · 1 build glm4-moe-tiny Text gen · MoE543M 131,072 tokens Apache 2.0 ~2.0 GiB2026-02-18
llama-nv-embed-reasoning NVIDIA · 1 build llama-nv-embed-reasoning-3b Embeddings3.2B 131,072 tokens Open ~7.6 GiB2026-02-18
minimax-m2-tiny PrimeIntellect · 1 build minimax-m2-tiny Text gen252M 4,096 tokens Apache 2.0 ~1.2 GiB2026-02-18
orange-nomic-1536 Orange · 1 build orange-nomic-v1.5-1536 Embeddings137M 8,192 tokens Apache 2.0 ~1.4 GiB2026-02-18
MiniMax-M2.5 PrimeIntellect · 1 build MiniMax-M2.5-bf16 Text gen228.7B 196,608 tokens Licence fee ~465.8 GiB2026-02-18
qwen3-moe-tiny PrimeIntellect · 1 build qwen3-moe-tiny Text gen · MoE670M 32,768 tokens Apache 2.0 ~2.3 GiB2026-02-18
embeddinggemma Red Hat · 1 build embeddinggemma-300m Embeddings303M 2,048 tokens Conditions ~4.3 GiB2026-02-18
nomic-embed-text Red Hat · 1 build nomic-embed-text-v1.5 Embeddings137M 8,192 tokens Apache 2.0 ~4.0 GiB2026-02-18
snowflake-arctic-embed-l Red Hat · 1 build snowflake-arctic-embed-l-v2.0 Embeddings568M 8,194 tokens Apache 2.0 ~2.2 GiB2026-02-18
ms_snsd_tfgridnet espnet · 1 build ms_snsd_tfgridnet Audio→audio ~1.6 GiB2026-02-16
NextStep-1.1-Pretrain-256px stepfun-ai · 1 build NextStep-1.1-Pretrain-256px Text→image15.0B 131,072 tokens Apache 2.0 ~35.9 GiB2026-02-16

Release dates use the upstream repository's publication date. The default is the past 12 months, through today. All time includes unknown and future dates. 3 families have no recorded date.

Size, context and GPU memory describe the listed build. For language models, the memory estimate covers one request at 8K context (or the model's shorter limit); image estimates cover one image. Other types use their existing weight and overhead estimates. Blank values are missing or incomplete and do not match numeric limits. Open a model to compare its builds and hardware estimates.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms