harn-vm 0.10.21

Async bytecode virtual machine for the Harn programming language
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

# Hugging Face Inference Providers router rows. The router is
# OpenAI-compatible, but availability and pricing can vary by selected
# upstream provider, so these rows encode stable model identity/capability
# metadata rather than a route-specific rate card.

[models."Qwen/Qwen3-Coder-480B-A35B-Instruct"]
name = "Qwen3-Coder 480B A35B Instruct (Hugging Face Router)"
provider = "huggingface"
wire_model = "Qwen/Qwen3-Coder-480B-A35B-Instruct"
context_window = 262144
logical_model = "qwen3-coder-480b-a35b"
equivalence_group = "qwen3-coder-480b-a35b"
served_variant = "huggingface-router"
api_dialect = "openai_chat"
tier = "frontier"
open_weight = true
strengths = ["coding", "agentic", "tool_use", "long_context"]
complementary_with = ["anthropic-claude", "openai-gpt", "google-gemini", "deepseek", "kimi"]
architecture = { parameter_count_b = 480.0, active_parameter_count_b = 35.0, moe = true, license = "Apache-2.0", source_url = "https://huggingface.co/Qwen/Qwen3-Coder-480B-A35B-Instruct", last_verified = "2026-07-15" }