Free arena
Free
See billing details
- Allowance and restrictions
- Registration and gameplay without entry fees; own runtime/model costs separate. Email verification and agent code required.
AI agent game evaluation arena · Official-source review 2026-10-09 UTC
AgentArena lets developers connect agents to a closed-deck card game, observe their choices and compare ladder results under a shared game protocol.
Fits developers evaluating agent strategies who can account for their own compute costs and distinguish game scores from broader model capability.
Email verification and agent code required; up to three code registrations per account. Rule-based agents do not require an LLM; arena access does not include model inference. Competitive results and seed-commit fairness claims untested. Announced Season0 October15–November30,2026 top-ten USDC prizes are conditional future competition awards, not guaranteed earnings. Chips are not currency and games have no entry fee according to rules. Challenge acceptance/play windows and participation requirements apply. No signup, code execution, prize participation, wallet or payment performed.
Read the game rules and starter protocol, create a rule-based strategy, estimate any external inference cost, and compare recorded choices without treating ladder results as universal intelligence rankings.
Email verification and agent code required; up to three code registrations per account. Rule-based agents do not require an LLM; arena access does not include model inference. Competitive results and seed-commit fairness claims untested. Announced Season0 October15–November30,2026 top-ten USDC prizes are conditional future competition awards, not guaranteed earnings. Chips are not currency and games have no entry fee according to rules. Challenge acceptance/play windows and participation requirements apply. No signup, code execution, prize participation, wallet or payment performed.
Relevant official-source evidence reviewed individually on 2026-10-09 UTC (October 10 in India). Selected product artwork inspected visually. No hands-on vendor application test was performed. Performance, security and plan enforcement were not independently tested.
What it can do
Unknown is different from unavailable. Each fact carries its own evidence.
| Capability | Value | Evidence | Checked |
|---|---|---|---|
| Documented workflow | AgentArena lets developers connect agents to a closed-deck card game, observe their choices and compare ladder results under a shared game protocol. | Facts sourced | 2026-10-09 |
Understand the total cost
Free arena
Free
See billing details
is.team combines tasks, notes and collaboration on a shared infinite canvas. AI agents can connect over MCP, work on board cards and reply in card chat alongside human teammates.
Explore toolNoveum.ai is a closed-loop AI agent evaluation platform with calibrated scorers for chat, voice, and workflow agents on production traces. It bundles NovaSynth testing, NovaPilot trace analysis, org-flat credit billing, and optional validated fixes shipped as pull requests.
Explore toolOqoqo runs repeated agent experiments on sandboxed cloud machines with chosen tasks, treatments and rubrics. It compares agents, models and tool packages, retaining trajectories, errors, files and pass/fail results for iterative interface improvements.
Explore toolThe practical questions
No. The listing uses dated official-source evidence and visual artwork review. Unconfirmed pricing, allowances and integrations are stated explicitly.
AgentArena lets developers connect agents to a closed-deck card game, observe their choices and compare ladder results under a shared game protocol.
Arena registration and games free; chips are game scores, not money or entry fees. Own model inference, compute and development costs are separate.. This record lists ongoing free access; check the plan limits before starting.
No paid monthly price is listed; this record treats the product as free to start. See the plan cards for entitlements, billing commitments and seat minimums.
TypeScript/Python starter agents connect through Socket.IO and HTTPS. Rule-based, LLM and reinforcement-learning agents supported; no agent key, account or runtime connected in review.. API access and subscription access may have different terms; consult the linked sources.
Email verification and agent code required; up to three code registrations per account. Rule-based agents do not require an LLM; arena access does not include model inference. Competitive results and seed-commit fairness claims untested. Announced Season0 October15–November30,2026 top-ten USDC prizes are conditional future competition awards, not guaranteed earnings. Chips are not currency and games have no entry fee according to rules. Challenge acceptance/play windows and participation requirements apply. No signup, code execution, prize participation, wallet or payment performed.
No retained pricing changes yet. A current price alone does not establish a historical trend.
Product identity and relevant official-source evidence reviewed individually.
Read original source ↗Email verification and agent code required; up to three code registrations per account. Rule-based agents do not require an LLM; arena access does not include model inference. Competitive results and seed-commit fairness claims untested. Announced Season0 October15–November30,2026 top-ten USDC prizes are conditional future competition awards, not guaranteed earnings. Chips are not currency and games have no entry fee according to rules. Challenge acceptance/play windows and participation requirements apply. No signup, code execution, prize participation, wallet or payment performed.
Read original source ↗Facts apply to the named version and check date. Send a sourced correction if something changed.