Arcee AI: Trinity Large Thinking (free)
Trinity Large Thinking is a powerful open source reasoning model from the team at Arcee AI. It shows strong performance in PinchBench, agentic workloads, and reasoning tasks. Launch video: https://youtu.be/Gc82AXLa0Rg?si=4RLn6WBz33qT--B7...
Anyone in the Space can @-mention Arcee AI: Trinity Large Thinking (free) with the team's shared context — pooled credits, one chat, one memory.
Starter is free forever — 1 Space, 100 credits/month, 1 MCP. No card.
Verdict
Best for
- Prototyping long-context applications at zero cost
- High-volume experimentation without budget limits
- Document analysis when benchmarks aren't critical
- Internal tooling with manual output validation
Strengths
The 262K context window handles full-length reports, codebases, and multi-document workflows without chunking. Zero-dollar pricing removes cost as a constraint for iteration, A/B testing, and exploratory work. Free tier with this much context is rare — most providers cap free inference at 8K-32K tokens or throttle heavily. Arcee's focus on efficient architectures suggests the model may punch above its weight class once benchmarks arrive.
Trade-offs
No public benchmark data means you're flying blind on accuracy, reasoning depth, and instruction-following relative to Claude, GPT-4, or Gemini. Free inference often comes with rate limits or deprioritized queue placement, though specifics aren't published. The 'Thinking' label implies chain-of-thought or reasoning focus, but without evals we can't confirm it matches o1-preview or DeepSeek's reasoning performance. If your use case demands proven accuracy, this model is too risky until benchmarks surface.
Specifications
- Provider
- arcee-ai
- Category
- llm
- Context length
- 262,144 tokens
- Max output
- 80,000 tokens
- Modalities
- text
- License
- proprietary
- Released
- 2026-04-01
Pricing
- Input
- $0.00/Mtok
- Output
- $0.00/Mtok
- Model ID
arcee-ai/trinity-large-thinking:free
Per-token prices show what the model costs upstream. On Switchy your team draws from one shared org credit pool — one plan, one balance for everyone.
Team cost calculator
5 seats · 80 msgs/day
Switchy meters this against your org's shared credit pool — one plan, one balance for everyone.
Providers
Performance
Benchmarks
Works well with
Top MCPs
Compatibility data comes from first-party telemetry; once we have enough co-usage signal, top MCPs for this model will appear here.
How Switchy teams use it
Starter prompts
Summarize Long Research Paper
Read the attached 40-page research paper and produce a 300-word summary covering the core hypothesis, methodology, key findings, and limitations. Prioritize clarity for a non-specialist audience.Open in a Space →
Extract Entities from Contract
Extract all party names, dates, monetary amounts, and termination clauses from this 80-page contract. Return results as a JSON object with keys: parties, dates, amounts, termination_clauses.Open in a Space →
Compare Three Product Specs
I've pasted three product specification documents below. Compare them on: feature completeness, pricing model, integration complexity, and support SLA. Rank them and justify your ranking in two paragraphs.Open in a Space →
Debug Multi-File Codebase
I've included five Python files from a Flask API. There's a bug causing 500 errors on POST /users. Trace the request flow, identify the root cause, and suggest a fix with line numbers.Open in a Space →
Generate FAQ from Transcript
Below is a 90-minute customer support call transcript. Generate a 10-question FAQ that covers the most common issues raised, with concise answers for each. Use the customer's own language where possible.Open in a Space →