Models & Endpoints.
Understand technical specifications, context limitations, and real-world use cases across Hyperedge capabilities and deployments.
Language Models (LLM)
Optimized text generation, coding, and reasoning models.
GPT OSS 20B
gpt-oss-20b
High-throughput 20B parameter model balanced for broad instruction-following and coding tasks.
GPT OSS Safeguard 20B
gpt-oss-safeguard-20b
Safety-aligned 20B model ideal for production and enterprise environments that require strict guardrails.
GPT OSS 120B
gpt-oss-120b
Massive 120B model geared towards deep reasoning, mathematics, and complex multi-step coding scenarios.
Llama 4 Scout (17Bx16E)
llama-4-scout
Mixture of Experts architecture balancing high accuracy with speed. Only activates essential parameters per token.
Qwen3 32B
qwen3-32b
Versatile dense model with exceptional multilingual abilities and leading benchmarks in coding capabilities.
Llama 3.3 70B Versatile
llama-3.3-70b-versatile
Powerful 70B general-purpose model providing near state-of-the-art text generation across all disciplines.
Llama 3.1 8B Instant
llama-3.1-8b-instant
Extremely fast 8B model optimized for real-time interactions, edge cases, and basic summarization.
Speech Models (Audio)
State-of-the-art TTS (Text-to-Speech) and ASR (Automatic Speech Recognition).
Canopy Labs Orpheus English
canopy-orpheus-en
High-fidelity, natural-sounding English voice generation prioritizing cadence and emotive inflection. Perfect for virtual agents.
Canopy Labs Orpheus Arabic Saudi
canopy-orpheus-ar
High-quality Arabic voice generation tailored specifically with Saudi dialects and localized pronunciation rules.
Whisper V3 Large
whisper-v3-large
High-accuracy automatic speech recognition with deep robustness against background noise and strong multilingual translation capabilities.
Whisper Large v3 Turbo
whisper-v3-turbo
Turbo-charged version of the V3 architecture optimized explicitly to vastly reduce the time-to-first-token in streaming transcriptions.
Built-in API Tools
Action-oriented endpoints that give LLMs connectivity, context, and computation.
Basic Search
General web search queries powered by high-speed indexers for real-time augmentation.
web_search
Advanced Search
Aggressive deep-page indexing that extracts greater textual context directly into the prompt stream.
web_search (advanced=true)
Visit Website
Directly fetch, parse, and strip HTML out of a specific target URL into clean markdown.
visit_website
Code Execution
Secure, isolated, ephemeral Python sandbox environments to evaluate mathematical formulas and logic.
code_interpreter
Browser Automation
Headless Chromium scripting capability triggered directly via the LLM to navigate SPAs and JS sites.
browser_automation
Ready to calculate the costs for your architectural deployment?
Use the Compute Calculator