Team Token Plans with 5x the usage.
One plan for your engineering team’s coding agents, in Claude Code, Codex and OpenCode. From $1,000 a month, on open models built for long runs.
Pi
Factory Droid
OpenHandsOne plan for the whole team.
Each plan’s daily allowance is worth 5x its price at usage-based rates.
Base
For teams putting agents into production
$1,000/ month
Get Base- $5,000+ of usage a month
- 100 concurrent requests
TeamMost popular
For engineering teams running agents all day
$5,000/ month
Get Team- $25,000+ of usage a month
- 500 concurrent requests
Heavy
For organizations running agents at scale
$10,000/ month
Get Heavy- $50,000+ of usage a month
- 1,000 concurrent requests
- Per-member API keys
- Usage dashboard
- Direct line to our engineers
Custom
For larger budgets and custom terms.
Volume pricing
Talk to an engineer- Higher rate limits
- Dedicated capacity
- Every model, shared across your team
- Resets daily at midnight UTC
- Past the allowance, requests bill per token
- Switch tiers or cancel any time
Usage is a plan’s daily allowance at our per-token rates over a 30-day month.
5x the usage for the same budget
What a month of Team plan usage costs without the plan.
Subconscious Team plan
$5,000Flat monthly fee, shared across your team
Subconscious, pay per token
$25,787Billed per request
GLM-5.3 on other platforms
$51,574Identical list price, but without context compression
Claude Opus 5.5
$56,952Frontier list price
Assumes agentic coding traffic: 98.2% cached input, 1.6% uncached input, 0.2% output. Rows without context compression carry 2x the tokens for the same work, because the harness re-sends a conversation that grows with every step. Our GLM-5.3 rate is the same list price Z.ai publishes, so that row is the cost of compression alone, not a cheaper model.
Three commands to your first run.
Install the CLI, sign in, and launch the agent you already use. Cloud agents take a plan key at an OpenAI- or Anthropic-compatible endpoint; see the setup guides.
# Install and sign in
$ npm install -g subconscious-cli
$ subc login
# Run OpenCode in Marathon Mode
$ subc opencode
●OpenCode now runs on subconscious/glm-5.3-marathon
Open models built for long coding runs.
Measured against the same open models on standard inference, at 200k+ tokens of context.
- Faster task completion
2x
- Measured at 200k+ tokens of context, same model and hardware.
- Lower cost
50%+
- The runtime compresses spans before they are billed, so you pay for what it processes.
- Higher accuracy
10%
- On agentic benchmarks. Less stale context gives the model a cleaner view of the task.
- Effective context
5M+
- Runs keep their history past the model’s window, without lossy summaries.
Questions
- A coding plan for engineering teams. You pay a flat monthly price and get a daily allowance worth 5x that price at usage-based rates, shared across your team and every model. Plans start at $1,000 a month.
- For teams that run agents every day, yes. The Team plan costs $5,000 a month and covers $25,000+ of usage at our per-token rates. Past the day’s allowance, requests move to usage-based pricing, so agents never stop mid-run.
- Yes. The subc claude command launches Claude Code on GLM-5.3 Marathon through our Anthropic-compatible endpoint. Codex, OpenCode and Pi have their own subc commands, and Copilot connects in VS Code. Your harness and workflow stay the same.
- Each plan has a daily allowance that resets at midnight UTC, rate limits, and a cap of 100, 500 or 1,000 concurrent requests by tier. Past the day’s allowance, requests move to usage-based pricing until the next reset. Enterprise plans get higher limits.
- Yes. One budget covers your team and every model. Heavy adds per-member API keys and a shared usage dashboard.
- No. Prompts and completions are never used to train anyone’s model, and the API keeps only token counts for billing.
- Yes. Switch tiers or cancel at any time.
Put your team’s agents in Marathon Mode.
Three commands, the harness you already use, and one allowance for the whole team.