Compare what the monthly price gets you: models, usage and ways to run them.
6,000 model calls per 5h, 45,000 weekly and 90,000 monthly.
Plan key and endpoint for interactive coding tools.
Pro availability limited daily; Lite discontinued. Not for unattended backend/script workloads.
Pro availability limited daily; Lite discontinued. Not for unattended backend/script workloads. One user prompt can trigger multiple billable model calls.
$20 monthly usage across models, Context Engine and compute; up to 50 seats.
CLI, Cosmos, Context Engine and MCP/native tools; pooled workspace usage.
Usage includes LLM, Context Engine and compute. LLM service fee 40%; no compute fee. Up to 50 seats included; dollars are not raw model-dollar equivalents.
$100 monthly usage across models, Context Engine and compute; up to 50 seats.
CLI, Cosmos, Context Engine and MCP/native tools; pooled workspace usage.
Usage includes LLM, Context Engine and compute. LLM service fee 40%; no compute fee. Up to 50 seats included; dollars are not raw model-dollar equivalents.
Lightweight desktop usage; numerical allowance unpublished.
ChatGPT sign-in: Codex app, CLI, IDE and cloud on Plus/Pro. Separate API-key billing is not included.
GPT-6 Luna
Codex and ChatGPT Work share allowance. Usage depends on model, context and task; no fixed token entitlement.
Price checked Oct 17 · Plan details checked Oct 16
Estimated local messages per 5h: Astra 5–45; Sol 15–160; Luna 350–3,000. Local and cloud share usage; weekly limits may apply.
ChatGPT sign-in: Codex app, CLI, IDE and cloud on Plus/Pro. Separate API-key billing is not included.
GPT-6.1 Sol, GPT-6 Luna; pricing table also lists GPT-6 Astra. Go: Luna Standard desktop, rollout dependent.
Codex and ChatGPT Work share allowance. Usage depends on model, context and task; no fixed token entitlement.
Price checked Oct 17 · Plan details checked Oct 16
No five-hour limit currently; total included allowance unpublished.
ChatGPT sign-in: Codex app, CLI, IDE and cloud on Plus/Pro. Separate API-key billing is not included.
GPT-6.1 Sol, GPT-6 Luna; pricing table also lists GPT-6 Astra. Go: Luna Standard desktop, rollout dependent.
Codex and ChatGPT Work share allowance. Usage depends on model, context and task; no fixed token entitlement.
Price checked Oct 17 · Plan details checked Oct 16
No five-hour limit currently; total included allowance unpublished.
ChatGPT sign-in: Codex app, CLI, IDE and cloud on Plus/Pro. Separate API-key billing is not included.
GPT-6.1 Sol, GPT-6 Luna; pricing table also lists GPT-6 Astra. Go: Luna Standard desktop, rollout dependent.
Codex and ChatGPT Work share allowance. Usage depends on model, context and task; no fixed token entitlement.
Price checked Oct 17 · Plan details checked Oct 16
No five-hour limit currently; total included allowance unpublished. Ultrafast access.
ChatGPT sign-in: Codex app, CLI, IDE and cloud on Plus/Pro. Separate API-key billing is not included.
GPT-6.1 Sol, GPT-6 Luna; pricing table also lists GPT-6 Astra. Go: Luna Standard desktop, rollout dependent.
Codex and ChatGPT Work share allowance. Usage depends on model, context and task; no fixed token entitlement.
Price checked Oct 17 · Plan details checked Oct 16
Daily request allowance, plus 6% off pay-as-you-go rates beyond it.
API and chat
Daily quota size is not published on this page.
Daily request allowance, plus 10% off pay-as-you-go rates beyond it.
API and chat
Daily quota size is not published on this page.
Base Pro usage; five-hour and weekly limits. Absolute quota unpublished.
Claude app and Claude Code share subscription usage.
Sonnet, Opus and Haiku; Fable through paid usage credits.
Five-hour and weekly limits; Fable requires paid usage credits. API billing is separate on Pro.
Price checked Oct 17 · Plan details checked Oct 16
5× Pro session usage, plus $100 Platform API credits per billing cycle.
Claude app and Code share subscription usage. Max additionally grants separate Claude Platform API credits.
Sonnet, Opus and Haiku; Fable capped at 50% of weekly usage.
5-hour and weekly app limits. API grant requires seven active days and linking one Console organization; expires each billing cycle. API grant excludes interactive Code and app extra usage; covers API and self-run Agent SDK.
Price checked Oct 17 · Plan details checked Oct 16
20× Pro session usage, plus $200 Platform API credits per billing cycle.
Claude app and Code share subscription usage. Max additionally grants separate Claude Platform API credits.
Sonnet, Opus and Haiku; Fable capped at 50% of weekly usage.
5-hour and weekly app limits. API grant requires seven active days and linking one Console organization; expires each billing cycle. API grant excludes interactive Code and app extra usage; covers API and self-run Agent SDK.
Price checked Oct 17 · Plan details checked Oct 16
Rolling five-hour, calendar-week and calendar-month limits; quantities unpublished.
Cline IDE/CLI and OpenAI-compatible API for scripts, apps and automation.
Reference API rates weight subscription usage. Processing fees may apply.
1× Plus weekly usage, shared with Grok; absolute quota unpublished.
Editor and Grok share a weekly usage pool. Premium+ cloud agents and reviews.
Grok API usage is billed separately at standard rates.
4× Plus weekly usage, shared with Grok; absolute quota unpublished.
Editor and Grok share a weekly usage pool. Premium+ cloud agents and reviews.
Grok API usage is billed separately at standard rates.
15× Plus weekly usage, shared with Grok; absolute quota unpublished.
Editor and Grok share a weekly usage pool. Premium+ cloud agents and reviews.
Grok API usage is billed separately at standard rates.
50× Plus weekly usage, shared with Grok; absolute quota unpublished.
Editor and Grok share a weekly usage pool. Premium+ cloud agents and reviews.
Grok API usage is billed separately at standard rates.
Daily and weekly usage limits; quantities unpublished. Inline edits and Tab completions included.
Devin Desktop, CLI and cloud agents; extra usage at API pricing.
Daily and weekly quota amounts are unpublished. The SWE-2 free-use promotion ended October 16, 2026.
Higher daily and weekly usage than Pro; quantities unpublished. Inline edits and Tab completions included.
Devin Desktop, CLI and cloud agents; extra usage at API pricing.
Daily and weekly quota amounts are unpublished. The SWE-2 free-use promotion ended October 16, 2026.
$15 credits/month: $10 base plus $5 variable flex allotment.
Copilot IDE, CLI, GitHub agents and Apps; credits cover these features, not unrestricted provider APIs.
Credits are dollars, not requests. Flex allotments may change. Pro+ and Max add third-party coding-agent delegation.
$70 credits/month: $39 base plus $31 variable flex allotment.
Copilot IDE, CLI, GitHub agents and Apps; credits cover these features, not unrestricted provider APIs.
Credits are dollars, not requests. Flex allotments may change. Pro+ and Max add third-party coding-agent delegation.
$200 credits/month: $100 base plus $100 variable flex allotment.
Copilot IDE, CLI, GitHub agents and Apps; credits cover these features, not unrestricted provider APIs.
Credits are dollars, not requests. Flex allotments may change. Pro+ and Max add third-party coding-agent delegation.
$19 paid credits/month. Bonus: 10% in month one, 50% in month two, then up to 40%.
Paid inference balance across IDE, CLI, Cloud Agents and Gateway; compatible external agents.
Bonus unlocks after paid credits and expires monthly. Credit processing fee 5%; cloud compute is separate.
$49 paid credits/month. Bonus: 10% in month one, 50% in month two, then up to 40%.
Paid inference balance across IDE, CLI, Cloud Agents and Gateway; compatible external agents.
Bonus unlocks after paid credits and expires monthly. Credit processing fee 5%; cloud compute is separate.
$199 paid credits/month. Bonus: 10% in month one, 50% in month two, then up to 40%.
Paid inference balance across IDE, CLI, Cloud Agents and Gateway; compatible external agents.
Bonus unlocks after paid credits and expires monthly. Credit processing fee 5%; cloud compute is separate.
2× Go shared monthly credits; rolling five-hour Code limit. Absolute quota unpublished.
Kimi membership with Code login/API key; membership-wide shared quota, distinct from Open Platform PAYG.
K3 and Code.
Only current China CNY pricing established. New plans: rolling 5-hour Code limit plus monthly shared credits, no weekly limit. Legacy plans keep older rules.
4× Go shared monthly credits; rolling five-hour Code limit. Absolute quota unpublished.
Kimi membership with Code login/API key; membership-wide shared quota, distinct from Open Platform PAYG.
K3 and Code; 1M context and HighSpeed.
Only current China CNY pricing established. New plans: rolling 5-hour Code limit plus monthly shared credits, no weekly limit. Legacy plans keep older rules.
14× Go shared monthly credits; rolling five-hour Code limit. Absolute quota unpublished.
Kimi membership with Code login/API key; membership-wide shared quota, distinct from Open Platform PAYG.
K3 and Code; 1M context and HighSpeed.
Only current China CNY pricing established. New plans: rolling 5-hour Code limit plus monthly shared credits, no weekly limit. Legacy plans keep older rules.
1× Go usage; absolute quota unpublished. Five-hour and weekly limits.
MiniMax Code sign-in or Subscription Key in compatible tools.
Text models, e.g., M3.1 Flash Preview; Image and speech models
Individual interactive use. Subscription Key uses plan quota, then purchased credit packs; PAYG keys are separate.
3× Go usage; absolute quota unpublished. Five-hour and weekly limits. Video quota resets weekly.
MiniMax Code sign-in or Subscription Key in compatible tools.
Text models, e.g., M3.1 Flash Preview; Video models, e.g., H3; Image and speech models
Individual interactive use. Subscription Key uses plan quota, then purchased credit packs; PAYG keys are separate.
7.5× Go usage; absolute quota unpublished. Five-hour and weekly limits. Video quota resets weekly.
MiniMax Code sign-in or Subscription Key in compatible tools.
Text models, e.g., M3.1 Flash Preview; Video models, e.g., H3; Image and speech models
Individual interactive use. Subscription Key uses plan quota, then purchased credit packs; PAYG keys are separate.
60 million included input tokens per week and 100 images per day.
Web and API
Included models use different allowance weights.
Model-specific coding allowances; Step 5 Preview has a temporary $60 monthly usage allowance.
Coding agents
Single-member workspace with coding-session identity.
Larger coding allowances; Step 5 Preview has a temporary $240 monthly usage allowance.
Coding agents
Single-member workspace with coding-session identity.
500 weighted requests per 5h plus $24 weekly credits; both limits apply. One concurrent request per model.
UI and API; any agent framework.
Both quota limits apply, regenerate incrementally, and scale by packs. Requests weighted relative to Kimi K3; embeddings do not use request quota.
2,000 weighted credits per 5h and 10,000 weekly; both limits apply.
Plan API endpoint for officially supported agent tools; not unrestricted general API.
Token-weighted credits; off-peak consumption halves. Peak Mon–Fri 14:00–18:00 Singapore time. Concurrency dynamically controlled.