Unfragile

Llama-3.3-SuperSwallow-70B-Instruct-v0.1 is a 70B-parameter Llama-family model, and none of the candidate sources names it directly. The closest matches are hosted or local runtimes that serve 70B Llama instruct models: Replicate hosts Meta's Llama 3 70B Instruct 2, Ollama runs open models locally or in the cloud 5, and LiteLLM can front many model endpoints behind one OpenAI-compatible key 6. TokenAssemble can check whether a 70B model fits a given GPU or Mac before you download it 10.

A question back: Do you want to run this model locally on your own hardware, or call it through a hosted API? If local, which GPU or Mac do you have?

1
Replicatereplicate.comverified and signed
Runs open-source models such as Meta's 70-billion-parameter Llama instruct model through a cloud API 2.
Website Docs Pricing Get started
+14 community tools built for it
for agents: meta/meta-llama-3-70b-instruct · lucataco/upstage-llama-2-70b-instruct-v2 · meta/codellama-70b-instruct
2
Ollamaollama.comwebsite read, tools not yet checked
Runs open AI models locally or in the cloud on macOS, Windows and Linux 5.
Website Docs Pricing Get started
+141 community tools built for it
3
LiteLLMlitellm.aifound in code, not yet checked
Open-source gateway that routes requests to 1,892 models across 140+ providers behind one OpenAI-compatible key 6.
Website Docs Pricing
+8 community tools built for it
for agents: shell · write_report · curl
4
TokenAssembletokenassemble.comwebsite read, tools not yet checked
Checks whether a GPU, Mac or mini-PC has enough VRAM or unified memory to run a local model, with quantization and context guidance 10.
Website Docs

Sources

2Replicate - Run AI with an APIreplicate.com · read 2026-10-29
5Ollama · Run AI models locally and in the cloudollama.com · read 2026-10-29
6LiteLLM — Open-Source AI Gateway & LLM Proxylitellm.ai · read 2026-10-29
10Local AI Hardware — GPU, VRAM & Model Fit | TokenAssembletokenassemble.com · read 2026-10-29

For agents: the 10 actions behind this answer

llama3-70b-instruct · Meta/Llama3-70b-instruct verified and signed
Task: Llama3-70b-instruct
meta/meta-llama-3-70b-instruct · meta/meta-llama-3-70b-instruct verified and signed
A 70 billion parameter language model from Meta, fine tuned for chat completions
llama.cpp · gguf verified and signed
· Llama website read, tools not yet checked
Run the latest open models on your computer. Chat with them, or connect them to your coding agents, editors, and apps. Free, private, and nothing to configure.
· 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.
shell · shin-watcher-oss found in code, not yet checked
Run a shell command in the cloned BerriAI/litellm working tree. Use this for git, uv, curl, ls, cat, pytest, etc.
· SGLang website read, tools not yet checked
SGLang powers fast, scalable inference for large language and multimodal models. Open-source serving framework with state-of-the-art performance.
ui_run · frictionless found in code, not yet checked
Execute Lua code in a session context
workers-ai-get-model-schema · Cloudflare API verified and signed
Get an AI model's input and output schemas
· TokenAssemble website read, tools not yet checked
Check if a GPU, Mac, or mini-PC can run a local AI model with Ollama, LM Studio or llama.cpp. VRAM fit, quantization, context length and speed tiers.

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.