Weekly credits
QwenCloud Token Plan $8 / $25 / $80 / month
Limited-time prices: $6 / $18 / $68 per month
A low entry price if Qwen is the model you want.
Qwen
works with Codex, Qwen Code, Claude Code, Cursor, OpenClaw
allowance, limits & sources included usage Lite / Standard / Pro: 2,500 / 10,000 / 40,000 credits per 7 days. before you pay Text, media and hosted tools share credits. Separate product from Alibaba Cloud Coding Plan. setup Dedicated Token Plan API key simultaneous use Advertised agent counts: 1–2 / 3–4 / 6–8 availability International; check country and account eligibility
Per-model allowance
OpenCode Go $10 / month
Several model families in one subscription.
GLM Kimi MiniMax Qwen DeepSeek MiMo GPT Grok Muse
works with OpenCode, Compatible API tools
allowance, limits & sources included usage Monthly allowances vary by model. Each also has a 5-hour cap of 20% and weekly cap of 50% of its monthly allowance. before you pay The model you choose changes how far the plan goes. Paid balance fallback is optional. setup Go API key simultaneous use Check current provider limits availability International; check country and account eligibility
Per-model allowance
ClinePass $9.99 / month
An open-model bundle you can also use outside Cline.
GLM Kimi DeepSeek MiMo MiniMax Qwen
works with Cline, Compatible API tools
allowance, limits & sources included usage Included usage varies by model; the provider advertises 2–5× standard API usage. before you pay Only ClinePass model IDs draw from this plan. Check the live allowance for your preferred model. setup Cline API key and ClinePass model ID simultaneous use Not confirmed availability International; check country and account eligibility
Per-model allowance
CommandCode $10 / $20 / $100 / $200 / month
Another multi-model option; compare the allowance for the exact model.
GPT GLM Kimi DeepSeek MiniMax Qwen MiMo Grok
works with Compatible API tools
allowance, limits & sources included usage GOAT starts at $10 with advertised $70 credits, subject to model-specific allowances. Higher tiers add usage and model access. before you pay Processing fees are extra. The $1 Go plan has no API access and is excluded from the starting price. setup API key; GOAT or higher simultaneous use Upstream rate limits apply availability International; check country and account eligibility
Weekly credits
Z.ai GLM Coding Plan from $18 / month
Direct access if you mainly want GLM.
GLM
works with OpenCode, Cline, Claude Code
allowance, limits & sources included usage Lite / Pro / Max: 2,000 / 12,000 / 28,000 credits per 5 hours and 10,000 / 60,000 / 140,000 per week. before you pay Tokens consume credits at different rates. Off-peak usage is charged at half the standard credit rate. setup Dedicated coding-plan connection simultaneous use Not confirmed availability International; check country and account eligibility
Shared token budget
MiniMax Token Plan $20 / $50 / $120 / month
Direct MiniMax access with larger tiers for heavier use.
MiniMax
works with Claude Code, Cline, OpenClaw, Compatible API tools
allowance, limits & sources included usage Plus / Max / Ultra share one allowance across supported text, image and speech models. before you pay Published token counts are estimates, not a promise of completed tasks. Media uses the same allowance. setup Token Plan key simultaneous use Advertised agents: 3–4 / 4–5 / 6–7 availability International; check country and account eligibility
Shared membership credits
Kimi Code check regional checkout / month
Worth checking if you already pay for Kimi.
Kimi
works with OpenCode, Roo Code, Claude Code, Hermes, OpenClaw
allowance, limits & sources included usage Coding is included in eligible Kimi memberships; several Kimi features use the same credit pool. before you pay International tier pricing was not confirmed. Use the coding-specific service, not the separate pay-as-you-go platform. setup Kimi Code membership key simultaneous use Tier and speed dependent availability International; check country and account eligibility
Monthly credits
Nous Portal / Hermes $20 / $100 / $200 / month
One balance for models and tools, with extra setup outside Hermes.
Hermes GPT Claude Gemini DeepSeek Qwen Kimi GLM MiniMax Grok
works with Hermes, Compatible API tools
allowance, limits & sources included usage Plus / Super / Ultra include $22 / $110 / $220 in monthly credits. Rollover caps: $10 / $50 / $100. before you pay Models, tools and cloud hosting spend the same balance. This is metered credit, not unlimited model access. setup Sign in to Hermes; local subscription proxy for other apps simultaneous use Check model/provider limits availability International; check country and account eligibility
Monthly credits
Ollama Cloud $20 / $100 / month
Hosted open models without buying hardware.
GLM Kimi MiniMax DeepSeek Qwen GPT-OSS
works with Codex, OpenCode, Claude Code, Hermes, Pi
allowance, limits & sources included usage Pro includes $60 of monthly usage; Max includes $300. before you pay Unused included credits do not roll over. Some models have peak-hour pricing. Local models are a separate, free option. setup Ollama cloud connection simultaneous use Max lists 10 concurrent requests availability International; check country and account eligibility
Weekly credits
Synthetic $30 per pack / month
A pack-based plan with explicit concurrency limits.
GLM Kimi Qwen GPT-OSS
works with Cline, Roo Code, Compatible API tools
allowance, limits & sources included usage Each pack: 500 requests per 5 hours and $24 of credits that regenerate over a week. before you pay One concurrent request per model per pack. Buying more packs increases the limits. setup API key simultaneous use 1 request per model per pack availability International; check country and account eligibility
Input-token allowance
NanoGPT check current checkout / month
Compare carefully if your agent sends a lot of context.
Multiple families
works with Compatible API tools
allowance, limits & sources included usage Pro advertises 60 million input-token units per week. Included models may have different multipliers. before you pay Cached input still counts. Explicit provider selection uses pay-as-you-go billing instead. Current tier price not confirmed. setup Subscription endpoint and included model simultaneous use Not confirmed availability International; check country and account eligibility
Request allowance
BytePlus Coding Plan $10 / $50 / month
A separate provider to consider alongside the usual model labs.
Seed
works with OpenCode, Cline, Roo Code, Claude Code, Cursor, Kilo
allowance, limits & sources included usage Lite and Pro subscriptions. Model availability and usage deductions depend on the current plan. before you pay Availability varies by country. Check the exact model list and current limits before buying. setup Coding Plan key simultaneous use Not confirmed availability International; check country and account eligibility
Credits
Xiaomi MiMo Token Plan check regional checkout / month
Direct MiMo access rather than buying it through a bundle.
MiMo
works with Kilo, Compatible API tools
allowance, limits & sources included usage Token Plan is available through regional connections; current tier prices and quotas need checkout verification. before you pay Use the correct regional Token Plan connection. Do not assume an older China-only price applies worldwide. setup Token Plan key; Europe, Singapore or China endpoint simultaneous use Not confirmed availability International; check country and account eligibility
Request allowance
Alibaba Cloud Coding Plan $50 / month
A multi-model plan with explicit request windows.
Qwen GLM Kimi MiniMax
works with OpenCode, Codex, Cline, Claude Code, Cursor, Hermes, Qwen Code, Kilo
allowance, limits & sources included usage Pro: 6,000 model calls per 5 hours, 45,000 per week and 90,000 per month. before you pay Limited subscription slots. Only listed model versions are included. Interactive coding use only; not an app-backend budget. setup Dedicated Coding Plan key and endpoint simultaneous use Additional dynamic limits apply availability International; check country and account eligibility
Monthly credits
GitHub Copilot from $10 / month
Existing subscribers may already have another route into OpenCode.
GPT Claude Other selected models
works with OpenCode, Copilot CLI, VS Code
allowance, limits & sources included usage Pro lists $15 total monthly AI credits. Pro+ and higher tiers offer more usage/model access. before you pay Unlimited completion does not mean unlimited agent use. Models and account policies affect availability. setup GitHub account sign-in simultaneous use Check account limits availability International; check country and account eligibility
Account allowance
SuperGrok / X Premium+ check eligible subscription / month
Check access you may already pay for through X.
Grok
works with Kilo, Hermes, Compatible API tools
allowance, limits & sources included usage Eligible subscriptions can sign in to Kilo; Hermes also documents an xAI subscription proxy. before you pay Current regional price and exact coding quota need checking. This does not include a fresh, separate allowance per tool. setup xAI sign-in; proxy for other compatible apps simultaneous use Not confirmed availability International; check country and account eligibility
Account allowance
ChatGPT / Codex $20 / $100 / $200 / month
The reference point: moving the same subscription will not fix an exhausted quota.
GPT
works with OpenCode, Codex, Kilo, Hermes
allowance, limits & sources included usage Paid tiers have different allowances. Pro $100 and $200 advertise 5× and 20× Plus usage. before you pay Changing coding tools shares the same account allowance. New $200 sign-ups/upgrades are paused; existing plans continue. setup ChatGPT account sign-in simultaneous use Subject to account/model limits availability International; check country and account eligibility
Subscription
StepFun Step Plan price to confirm / month
A further option that needs more verification.
StepFun
works with Compatibility to confirm
allowance, limits & sources included usage Official global and China Step Plan pages exist. Exact current paid allowances were not confirmed. before you pay Included for coverage, not a buying recommendation. Check region, supported clients and current paid tiers. setup Check official setup simultaneous use Not confirmed availability International; check country and account eligibility
Request allowance
Tencent Cloud Coding Plan ¥40 / ¥200 / month
Regional alternative with a multi-model catalog.
Hunyuan GLM Kimi MiniMax
works with OpenCode, Codex, Cline, Claude Code, Cursor, Kilo, OpenClaw
allowance, limits & sources included usage Lite: 1,200 calls per 5 hours / 9,000 weekly / 18,000 monthly. Pro: 6,000 / 45,000 / 90,000. before you pay CNY-priced regional offering. Account eligibility and purchase availability need checking. Not converted to USD for sorting. setup Dedicated Coding Plan key simultaneous use Check current provider limits availability China / regional; check country and account eligibility
Account / organization
GitLab Duo check GitLab plan / month
Check an existing work subscription before buying another personal plan.
Plan-dependent
works with OpenCode
allowance, limits & sources included usage OpenCode documents a GitLab Duo connection. Entitlement depends on the user, group or instance. before you pay Agent Platform setup and organization policies matter. Current standalone cost and allowance not confirmed. setup GitLab Duo and Agent Platform enabled simultaneous use Organization dependent availability International; check country and account eligibility BEFORE YOU SUBSCRIBE
same monthly price ≠ same amount of work credits aren't a common currency $60 of usage at one provider is not necessarily the same number of tokens at another. Model rates, cache discounts and tool charges change the result.
one task can make many calls An agent can read files, run tests and retry before it finishes one request. Published call counts are not counts of completed coding tasks.
check the exact model and tier Model tags show selected families, not every version or every tier. Sign-in shares your existing allowance. It does not reset it.
scope, sources & what is still missing This is a dated research snapshot of subscriptions usable through multiple coding tools, including regional and partially verified options. It is not a benchmark, an uptime test or a complete list of every reseller. “Documented” means published terms were located, not that I personally tested the plan.
Prices are monthly list prices unless marked as a promotion. Unknown and non-USD prices sort last and do not appear in budget-filtered results. Checkout is the final source for taxes, region, availability and changed terms.
Claude, Gemini and Muse consumer subscriptions are outside this comparison's portable-plan scope. Their models may still be available through another provider. Pay-as-you-go-only services are not counted as subscriptions.
Coverage gaps: smaller resellers such as Standard Compute and DevPass, and additional regional offerings, still need verification. Kimi, NanoGPT, MiMo, xAI, StepFun and GitLab have specific open questions marked on their cards. This page is manually maintained, not a live price feed.
found a change or a missing plan? send me the official source ↗
More things I'm testing: build · models drawing in pixels
