No product named MiniMax-H3-GGUF was found in the sources. The request names minimax-h3, which appears in source 1 only as Pruna's p-video-2-pro-reference, a model built on MiniMax H3 for video generation, not as a GGUF weight file. For running GGUF weights locally, the relevant sources are Hugging Face (GGUF model discovery and local running) and ggml.ai (the library behind GGUF), with Ollama as a local runner.
A question back: Do you want to run the MiniMax-H3 GGUF weights locally (and on what hardware), or call MiniMax H3 through a hosted API for video generation?
- 1
Replicatereplicate.comverified and signed
Hosts Pruna's prunaai/p-video-2-pro-reference, a reference-to-video model built on MiniMax H3 that generates 5–15 second video from a prompt plus images 9.
for agents: prunaai/p-video-2-pro-reference · prunaai/p-video-avatar · prunaai/p-video-2-pro
- 2
Hugging Facehuggingface.coverified and signed
Hosts open models and includes a huggingface-local-models action to find GGUF files and run them locally with llama.cpp on CPU or GPU 11.
official:MCP server · @huggingface/mcp-clientskill · hf-cliskill · hf-cloud-aws-context-discoveryskill · hf-cloud-python-env-setupskill · hf-cloud-sagemaker-deployment-planner
for agents: huggingface-local-models · hf-mem · huggingface-best
- 3
- 4
Ollamaollama.comwebsite read, tools not yet checked
Runs open AI models locally or in the cloud on macOS, Windows and Linux 6.
+137 community tools built for it
- 5
ArkRouteark-route.comwebsite read, tools not yet checked
Offers one OpenAI-compatible API for 31 models, including MiniMax H3, for hosted video and chat generation rather than local GGUF use 7.
Sources
- 1Replicate - Run AI with an APIreplicate.com · read 2026-10-29
- 3Hugging Face – The AI community building the future.huggingface.co · read 2026-10-29
- 6Ollama · Run AI models locally and in the cloudollama.com · read 2026-10-29
- 7ArkRoute — 31 AI Models, One APIark-route.com · read 2026-10-29
- 8ggml.aiggml.ai · read 2026-10-29
- 9prunaai/p-video-2-pro-reference: declares "prunaai/p-video-2-pro-reference" (the record we read and signed)unfragile.ai · read 2026-10-24
- 11huggingface-local-models: declares "huggingface-local-models" (the record we read and signed)unfragile.ai · read 2026-10-24
For agents: the 10 actions behind this answer
prunaai/p-video-2-pro-reference · prunaai/p-video-2-pro-reference verified and signed
Pruna's reference-to-video model, built on MiniMax H3. Generate 5–15 second video from a prompt plus up to 9 images, 3 videos, and 3 audio clips, with speed and cost modes.
nano-banana-pro-openrouter · nano-banana-pro-openrouter verified and signed
Generate or edit images via OpenRouter with the Gemini 3 Pro Image model. Use for prompt-only image generation, image edits, and multi-image compositing; supports 1K/2K/4K output.
huggingface-local-models · huggingface-local-models verified and signed
Use to select models to run locally with llama.cpp and GGUF on CPU, Mac Metal, CUDA, or ROCm. Covers finding GGUFs, quant selection, running servers, exact GGUF file lookup, conversion, and OpenAI-compatible local serving.
github · ghapi verified and signed
workers-ai-search-model · Cloudflare API verified and signed
Model Search
spuuntries/flatdolphinmaid-8x7b-gguf · spuuntries/flatdolphinmaid-8x7b-gguf verified and signed
Undi95's FlatDolphinMaid 8x7B Mixtral Merge, GGUF Q5_K_M quantized by TheBloke.
spuuntries/miqumaid-v1-70b-gguf · spuuntries/miqumaid-v1-70b-gguf verified and signed
NeverSleep's MiquMaid v1 70B Miqu Finetune, GGUF Q3_K_M quantized by NeverSleep.
· Ollama website read, tools not yet checked
Ollama is the easiest way to run and build with open AI models locally or in the cloud on macOS, Windows, and Linux.
mikeei/dolphin-2.9-llama3-70b-gguf · mikeei/dolphin-2.9-llama3-70b-gguf verified and signed
Dolphin is uncensored. I have filtered the dataset to remove alignment and bias. This makes the model more compliant.
minimax/speech-2.6-hd · minimax/speech-2.6-hd verified and signed
MiniMax Speech 2.6 HD delivers studio-quality multilingual text-to-audio on Replicate with nuanced prosody, subtitle export, and premium voices
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.