Loading Apertis coding plans
Coding Plans
Code More. Pay Less.
Claude Opus 4.8, GLM-5.2, DeepSeek V4 Pro, Kimi K2.6 — all in one plan with lower prices.
✓ 30+ AI Models · ✓ Priority Support · ✓ Cancel Anytime
Claude CodeYour Agent
Apertis Connected
Refactor auth module to support OAuth 2.0 with PKCE flow.
Apertis Connected
Build a REST API with pagination and rate limiting.
Apertis Connected
Review changes and keep local agent work aligned with the repo.
Apertis Connected
Add Stripe subscription billing to the backend.
Apertis Connected
Create a real-time notification system with WebSocket.
Apertis Connected
Optimize database queries for multi-tenant architecture.
Apertis Connected
Wire a new provider route without changing the OpenAI client.
Apertis Connected
Trace a bug, edit the fix, and run the verification loop.
Apertis Coding Plan
- Fixed Monthly Cost
- 30+ Premium Models
- Free Cache Hits
- Auto-Failover
- Priority Routing
- Unlimited Free Models
30+ Models
OpenAI
Anthropic
Google
MiniMax
xAI
Z.AI
Works anywhere
One key for your coding stack.
Use the same OpenAI-compatible endpoint across your favorite agents. Pick a plan, keep your tools, and let quota stay predictable.
30+ model routesFree cache hitsPAYG fallback
Predictable plans
Pricing that stays out of your flow.
Choose the amount of agent work you want each month. The cards stay simple; detailed model availability lives below.
Lite
For lightweight coding work
$33/ quarter
600 quotas / monthQuota cost varies by model multiplier. See rates in Available Models below.
Included
- Compatible with over 10 coding tools, including Claude Code, Roo Code, Cline, Kilo Code, OpenCode, Crush, and Goose, with more being continuously added
- Compatible for your OpenClaw (prev. Moltbot & Clawdbot)
- 600 quotas per month
Pro
For professional workloads
$68.25/ quarter
900 quotas / monthQuota cost varies by model multiplier. See rates in Available Models below.
Included
- Compatible with over 10 coding tools, including Claude Code, Roo Code, Cline, Kilo Code, OpenCode, Crush, and Goose, with more being continuously added
- 900 quotas per month
- Compatible for your OpenClaw (prev. Moltbot & Clawdbot)
Plus
For sustained agent sessions
$162/ quarter
1,500 quotas / monthQuota cost varies by model multiplier. See rates in Available Models below.
Included
- Compatible with over 10 coding tools, including Claude Code, Roo Code, Cline, Kilo Code, OpenCode, Crush, and Goose, with more being continuously added
- 1500 quotas per month
- Compatible for your OpenClaw (prev. Moltbot & Clawdbot)
Max
For high-volume automation
$540/ quarter
3,000 quotas / monthQuota cost varies by model multiplier. See rates in Available Models below.
Included
- Compatible with over 10 coding tools, including Claude Code, Roo Code, Cline, Kilo Code, OpenCode, Crush, and Goose, with more being continuously added
- 3000 quotas per month
- Compatible for your OpenClaw (prev. Moltbot & Clawdbot)
Available Models by Plan
Subscription plans use rate-based billing. Each request consumes quota based on the model's rate multiplier.Free ModelsUnlimited usesIncluded in all plans - no quota consumed
Nemotron 3 Nano Omni
MiniMax M2.1
Gemini 2.5 Flash
GPT-5.1 Codex (Mini)
GPT OSS 120B
DeepSeek V3.2
DeepSeek V3.1
Lite
50 modelsModelRate·Est. Uses
Qwen3.5-Flash
0.04x·15,000Gemma 4 31B
0.04x·15,000MiniMax M2
0.05x·12,000MiniMax M2.7
0.12x·5,000MiniMax M2.5
0.12x·5,000Gemini 3.1 Flash Lite Preview
0.12x·5,000Kimi K2 Thinking
0.14x·4,285Kimi K2.5
0.17x·3,529Gemini 3 Flash Preview
0.2x·3,000DeepSeek V4 Flash
0.21x·2,857Claude Haiku 4.5 (Claude Code)
0.21x·2,857Claude Haiku 4.5
0.21x·2,857GLM 5
0.24x·2,500GLM 5 Turbo
0.38x·1,578Qwen3.5-35B-A3B
0.4x·1,500Qwen3.5-27B
0.4x·1,500Qwen3.5-122B-A10B
0.5x·1,200Qwen3.5 Plus 2026-02-15
0.5x·1,200DeepSeek V4 Pro
0.5x·1,200GLM 4.6
0.59x·1,016MiMo-V2.5
0.6x·1,000Claude Sonnet 4.5 (Claude Code)
0.63x·952Qwen3.5 397B A17B
0.7x·857Gemini 3.1 Pro Preview
0.77x·779GPT-5.4
0.86x·697Qwen3.7 Plus
1.5x·400Qwen3.6 Plus
1.5x·400MiniMax M3
1.5x·400Kimi K2.7 Code
1.5x·400Kimi K2.6
1.5x·400GLM 5.2
1.5x·400GLM 5.1
1.5x·400Claude Sonnet 4.6 (Claude Code)
1.5x·400Claude Sonnet 4.6
1.5x·400Claude Opus 4.5 (Claude Code)
1.5x·400Claude Opus 4.5
1.5x·400Qwen3.7 Max
2x·300GPT-5.5
3x·200Claude Opus 4.8 (Claude Code)
5x·120Claude Opus 4.8
5x·120Claude Opus 4.7 (Thinking) (Claude Code)
5x·120Claude Opus 4.7 (Thinking)
5x·120Claude Opus 4.7 (Claude Code)
5x·120Claude Opus 4.7
5x·120Claude Opus 4.6 (Thinking) (Claude Code)
10x·60Claude Opus 4.6 (Thinking)
10x·60Claude Opus 4.6 (Claude Code)
10x·60Claude Opus 4.6
10x·60GPT-5.4 Pro
10.4x·57ProPopular
53 modelsModelRate·Est. Uses
Qwen3.5-Flash
0.02x·45,000Gemma 4 31B
0.02x·45,000MiniMax M2
0.04x·22,500MiniMax M2.7
0.08x·11,250MiniMax M2.5
0.08x·11,250Gemini 3 Flash Preview
0.08x·11,250GLM 4.7
0.08x·11,250Gemini 3.1 Flash Lite Preview
0.09x·10,000Kimi K2 Thinking
0.1x·9,000Kimi K2.5
0.12x·7,500DeepSeek V4 Flash
0.15x·6,000Claude Haiku 4.5 (Claude Code)
0.15x·6,000Claude Haiku 4.5
0.15x·6,000GLM 5
0.17x·5,294GLM 5 Turbo
0.27x·3,333Qwen3.5-35B-A3B
0.3x·3,000Qwen3.5-27B
0.3x·3,000Gemini 3.1 Pro Preview
0.31x·2,903DeepSeek V4 Pro
0.34x·2,647Qwen3.5-122B-A10B
0.4x·2,250Qwen3.5 Plus 2026-02-15
0.4x·2,250GLM 4.6
0.42x·2,142Claude Sonnet 4.5 (Claude Code)
0.45x·2,000Qwen3.5 397B A17B
0.5x·1,800GPT-5.1 Codex
0.54x·1,666MiMo-V2.5
0.6x·1,500GPT-5 Codex
0.91x·989Qwen3.7 Plus
1x·900Qwen3.6 Plus
1x·900MiniMax M3
1x·900Kimi K2.7 Code
1x·900Kimi K2.6
1x·900GPT-5.4
1x·900Claude Opus 4.5 (Claude Code)
1.1x·818Claude Opus 4.5
1.1x·818Claude Sonnet 4.6 (Claude Code)
1.3x·692Qwen3.7 Max
1.4x·642GLM 5.2
1.5x·600GLM 5.1
1.5x·600Claude Sonnet 4.6
1.5x·600GPT-5.5
1.75x·514Claude Opus 4.8 (Claude Code)
3.5x·257Claude Opus 4.8
3.5x·257Claude Opus 4.7 (Thinking) (Claude Code)
3.5x·257Claude Opus 4.7 (Thinking)
3.5x·257Claude Opus 4.7 (Claude Code)
3.5x·257Claude Opus 4.7
3.5x·257Claude Opus 4.6 (Thinking) (Claude Code)
3.5x·257Claude Opus 4.6 (Thinking)
3.5x·257Claude Opus 4.6 (Claude Code)
3.5x·257Claude Opus 4.6
3.5x·257GPT-5.4 Pro
7.4x·121Plus
59 modelsModelRate·Est. Uses
Qwen3.5-Flash
0.02x·75,000Gemma 4 31B
0.02x·75,000MiniMax M2
0.03x·50,000MiniMax M2.7
0.05x·30,000MiniMax M2.5
0.05x·30,000Gemini 3 Flash Preview
0.05x·30,000GLM 4.7
0.05x·30,000Gemini 3.1 Flash Lite Preview
0.06x·25,000Kimi K2 Thinking
0.07x·21,428Kimi K2.5
0.08x·18,750DeepSeek V4 Flash
0.1x·15,000Claude Haiku 4.5 (Claude Code)
0.1x·15,000Claude Haiku 4.5
0.1x·15,000GLM 5
0.12x·12,500GLM 5 Turbo
0.19x·7,894Gemini 3.1 Pro Preview
0.22x·6,818Qwen3.5-35B-A3B
0.25x·6,000Qwen3.5-27B
0.25x·6,000DeepSeek V4 Pro
0.27x·5,555Qwen3.5-122B-A10B
0.3x·5,000Qwen3.5 Plus 2026-02-15
0.3x·5,000MiMo-V2.5
0.3x·5,000
GPT-5.1-Codex-Max
0.3x·5,000GPT-5.1 Codex (High)
0.3x·5,000GLM 4.6
0.3x·5,000Claude Sonnet 4.5 (Claude Code)
0.3x·5,000GPT-5.1 Codex
0.38x·3,947Qwen3.5 397B A17B
0.4x·3,750GPT-5 Codex
0.65x·2,307Claude Opus 4.5 (Claude Code)
0.75x·2,000Claude Opus 4.5
0.75x·2,000Qwen3.7 Plus
0.8x·1,875Qwen3.6 Plus
0.8x·1,875MiniMax M3
0.8x·1,875Kimi K2.7 Code
0.8x·1,875Kimi K2.6
0.8x·1,875GPT-5.4
0.8x·1,875Claude Sonnet 4.6 (Claude Code)
0.8x·1,875Claude Sonnet 4.6
0.8x·1,875Qwen3.7 Max
1x·1,500GLM 5.2
1x·1,500GLM 5.1
1x·1,500GPT-5.5
1.5x·1,000Claude Opus 4.8 (Claude Code)
3x·500Claude Opus 4.8
3x·500Claude Opus 4.7 (Thinking) (Claude Code)
3x·500Claude Opus 4.7 (Thinking)
3x·500Claude Opus 4.7 (Claude Code)
3x·500Claude Opus 4.7
3x·500Claude Opus 4.6 (Thinking) (Claude Code)
3x·500Claude Opus 4.6 (Thinking)
3x·500Claude Opus 4.6 (Claude Code)
3x·500Claude Opus 4.6
3x·500Claude Opus 4.1 (Claude Code)
3.8x·394GPT-5.4 Pro
5x·300GPT-5 Pro
5.5x·272Max
59 modelsModelRate·Est. Uses
Qwen3.5-Flash
0.01x·300,000MiniMax M2
0.01x·300,000Gemma 4 31B
0.01x·300,000MiniMax M2.7
0.03x·100,000MiniMax M2.5
0.03x·100,000Gemini 3.1 Flash Lite Preview
0.03x·100,000Gemini 3 Flash Preview
0.03x·100,000GLM 4.7
0.03x·100,000Kimi K2 Thinking
0.04x·75,000Kimi K2.5
0.05x·60,000DeepSeek V4 Flash
0.06x·50,000Claude Haiku 4.5 (Claude Code)
0.06x·50,000Claude Haiku 4.5
0.06x·50,000GLM 5
0.07x·42,857GLM 5 Turbo
0.11x·27,272Gemini 3.1 Pro Preview
0.13x·23,076GLM 5.2
0.15x·20,000GLM 5.1
0.15x·20,000GLM 4.6
0.17x·17,647DeepSeek V4 Pro
0.17x·17,647Claude Sonnet 4.5 (Claude Code)
0.18x·16,666Qwen3.5-35B-A3B
0.2x·15,000Qwen3.5-27B
0.2x·15,000Qwen3.5-122B-A10B
0.2x·15,000Qwen3.5 Plus 2026-02-15
0.2x·15,000MiMo-V2.5
0.2x·15,000GPT-5.1-Codex-Max
0.22x·13,636GPT-5.1 Codex (High)
0.22x·13,636GPT-5.1 Codex
0.22x·13,636GPT-5.4
0.26x·11,538Qwen3.5 397B A17B
0.3x·10,000GPT-5 Codex
0.38x·7,894Claude Opus 4.5 (Claude Code)
0.47x·6,382Claude Opus 4.5
0.47x·6,382Qwen3.7 Plus
0.5x·6,000Qwen3.6 Plus
0.5x·6,000MiniMax M3
0.5x·6,000Kimi K2.7 Code
0.5x·6,000Kimi K2.6
0.5x·6,000Claude Sonnet 4.6 (Claude Code)
0.5x·6,000Claude Sonnet 4.6
0.5x·6,000Qwen3.7 Max
0.6x·5,000GPT-5.5
0.75x·4,000Claude Opus 4.8 (Claude Code)
2.5x·1,200Claude Opus 4.8
2.5x·1,200Claude Opus 4.7 (Thinking) (Claude Code)
2.5x·1,200Claude Opus 4.7 (Thinking)
2.5x·1,200Claude Opus 4.7 (Claude Code)
2.5x·1,200Claude Opus 4.7
2.5x·1,200Claude Opus 4.6 (Thinking) (Claude Code)
2.5x·1,200Claude Opus 4.6 (Thinking)
2.5x·1,200Claude Opus 4.6 (Claude Code)
2.5x·1,200Claude Opus 4.6
2.5x·1,200Claude Opus 4.1 (Claude Code)
2.9x·1,034GPT-5.4 Pro
3.1x·967GPT-5 Pro
4.2x·714Frequently Asked Questions
Everything you need to know about Apertis Coding Plans
Still have questions?
Check out our documentation→What happens when my quota runs out?
When your subscription quota is exhausted, you can enable Pay-As-You-Go (PAYG) as a backup. PAYG allows you to continue using the service by consuming your account balance at standard rates. You can set spending limits and alerts in your dashboard.
Is there a 5-hour quota window or a per-request token cap?
No. Apertis has no time-based quota window (no 5-hour reset) and no Apertis-imposed per-request token ceiling. Each request can use up to the underlying model's own context window - the limit is the model, not us. Your subscription quota is measured in monthly usage units, not clock time.
Can I use higher-tier models with On-Demand Usage?
No, On-Demand Usage (PAYG) respects your subscription tier limits. Lite users can only access Lite models, Pro users can access Pro + Lite models, and Max users can access all models. This ensures that subscription tiers maintain their value. If you need access to higher-tier models, please upgrade your subscription plan.
Can I upgrade or downgrade mid-cycle?
Yes! You can upgrade your plan anytime and start using higher-tier models immediately. The price difference is prorated for the remaining cycle. Downgrades take effect at the start of your next billing cycle to ensure you get full value from your current plan.
How does the dedicated token work?
Each subscription includes a dedicated API token that is automatically created and managed. This token has higher priority routing and guaranteed quota. The token is automatically deleted when your subscription expires, and you can renew it anytime in your dashboard.
What is the refund policy?
You can cancel your subscription anytime from your account Settings - you'll retain full access until your current billing period ends with no further charges. For first-time subscriptions, you may request a prorated refund within 3 days of payment by emailing [email protected]. Refunds are calculated based on quota usage. Plan upgrades are not eligible for the 3-day window - see the next FAQ for upgrade rules. Full Refund Policy: apertis.ai/refund.
How do plan upgrades and downgrades work?
Upgrades take effect immediately: you receive your new plan's full monthly quota, are charged the new plan's full monthly fee, and your billing cycle resets to the upgrade date. The previous plan's remaining time and any unused quota are forfeited and not refunded. Downgrades take effect at the end of your current billing cycle - you keep full access until then with no immediate charge or refund. For annual plans, the unused remainder is converted to a Stripe account credit applied to future invoices. Full details in our Terms of Service and Refund Policy.
Which coding tools are supported?
All plans support Claude Code, Roo Code, Kilo Code, Cline, OpenCode, Crush, and Goose. Additional tools are continuously being added. Check the "Supported Coding Tools" section below for the complete list.
How are invoices handled?
Invoices are generated automatically each billing cycle and sent to your email. You can view and download all past invoices from your subscription dashboard. We use Stripe for secure payment processing and invoice management.
Trusted by Developers Worldwide
See what developers are saying about building with Apertis
Swapped our base_url, ran the test suite. Everything passed first try. Two minutes from start to production. The OpenAI SDK compatibility is real, not marketing.
ML
Staff Engineer, Fintech
Coding Plan gave me a fixed monthly cost for Claude Code instead of watching tokens spike. Cache hits are free - my AI bill dropped ~35% without changing anything.
TH
Freelance Developer
OpenAI went down 20 minutes before our investor demo. Apertis rerouted to Claude automatically. Demo went perfectly - nobody noticed.
DO
Platform Lead, Healthcare SaaS
We send a lot of repeated prompts for product descriptions. Prompt cache cut our bill by about 40%. Cache reads are completely free, and latency dropped too.
RP
ML Engineer, E-commerce
Used to juggle separate API keys for OpenAI, Anthropic, and Google. Now it's one key, one dashboard. Onboarding new devs went from a full day to under an hour.
JN
Backend Engineer, Dev Agency
Integrated via the Vercel AI SDK provider package. Streaming, tool calls, structured output - it all just works. Best DX I've had with any AI gateway.
SB
Full-stack Developer, AI Startup
Ready to Supercharge Your Coding?
Join thousands of developers shipping faster with AI-powered coding assistance. No credit card required to start.