Comparing LLM API providers?
Each page compares one gateway, plan or API with Standard Compute and says plainly when the other one is the better pick. Standard Compute optimizes for cost: each request goes to the cheapest model that can do it, on providers we vet for price, uptime and speed, with automatic fallback. Europe or US rules are opt-in, and Standard Compute doesn't train on your prompts or store them.
Gateways and routers
- OpenRouterPay-per-token aggregator for hundreds of models, with provider routing and fallbacks.
- Atlas CloudChat, image, video and audio models behind one key, paid as you go.
- Router9Flat-price AI studio and API for agent harnesses, with monthly credit quotas and budgets per key.
- RequestyLLM gateway for hundreds of models at provider rates plus a top-up fee, with an EU endpoint.
- OpenCode ZenOpenCode's pay-per-token model catalog, now called OpenCode Console in its docs.
- nexos.aiEuropean AI gateway at provider prices plus a fee, with business controls and a chat workspace.
- PortkeyEnterprise AI gateway and control plane, now sold by Palo Alto Networks; priced by logs.
- Cloudflare AI GatewayFree gateway layer on Cloudflare, with paid Unified Billing credits and an Auto Router.
- NanoGPTHundreds of text and image models paid as you go, plus an open-weight subscription.
- CodeRouterSmart-routing API for coding agents with monthly plans counted in tokens.
- Together LinkTogether AI's free beta CLI that runs Claude Code, Codex and OpenCode on open models.
- OpperStockholm-based gateway for hundreds of models at provider rates plus a small credit fee.
- Vercel AI GatewayVercel's model gateway at provider list price with no markup, from prepaid credits.
- LiteLLMOpen-source, self-hosted AI gateway for many providers on your own keys.
Coding-agent plans and passes
- Standard CodeManaged cloud coding agent billed per concurrent session; a different company from us.
- GitHub CopilotGitHub's per-seat AI assistant inside the editor, with a monthly AI Credits allowance.
- DevPassLLM Gateway's coding subscription, which cut its included usage on October 15, 2026.
- OpenCode GoOpenCode's low-cost model subscription, with 5-hour, weekly and monthly limits per model.
- CursorThe AI code editor, sold in plans from Plus to Ultra with a usage pool that refreshes weekly.
- DevinCognition's coding agent and Devin Desktop, the editor formerly called Windsurf, on shared plans with daily and weekly quotas.
- Cline PassCline's subscription to a curated open-weight lineup, with an unstated usage quota.
- Nous PortalNous Research's Hermes Agent subscription: many models, managed tools and monthly credits.
- Kilo PassKilo's monthly credit subscription for Kilo Code, with bonus credits that grow with a streak.
Model vendors' own plans
- MiniMax M PlanMiniMax's subscription for its own M3.1 Flash Preview plus image, audio and video models, metered in 5-hour and weekly windows.
- GLM Coding Plan (Z.ai)Z.ai's GLM-5.3 subscription, metered in credit-based 5-hour and weekly windows.
- Qwen Coding Plan (Alibaba)Alibaba's multi-vendor open-weight coding plan, with every request window published.
- Claude MaxAnthropic's consumer plan with Claude Code, metered in 5-hour and weekly windows.
- Claude UltraAnthropic's top consumer plan, called Max 20x until October 2026, metered by a weekly limit once verified.
- ChatGPT Plus / ProOpenAI's consumer plans with the Codex agent, with usage limits that scale by tier.
- Kimi CodeMoonshot's coding agent and quota, included in Kimi memberships from Plus, with a rolling 5-hour window.
- Google AntigravityGoogle's agentic IDE and CLI, with quotas set by your Google AI plan and refreshed every five hours or weekly.
Flat-price open-model subscriptions
- Awan LLMOpen Llama models with daily request caps instead of token metering.
- SyntheticEight open-weight models metered by requests, with a no-training, no-storage stance.
- ChutesOpen models on a decentralized network, with a daily quota and a value cap.
- CheapestInferenceOpen-weight model pools inside reserved daily time blocks.
- FeatherlessThousands of open-weight models with model pinning and agent sandboxes.
Direct model APIs and inference clouds
- OpenAI APIOpenAI's own pay-per-token API for GPT models.
- Anthropic APIAnthropic's own pay-per-token API for Claude models.
- Together AIFast pay-per-token inference for open models, with fine-tuning and dedicated endpoints.
- GroqVery low-latency inference on custom LPU hardware for a small set of open models.
- Fireworks AIFast pay-per-token inference cloud for open models, with fine-tuning.
Plan pricing and limits explained
Side by side with Standard Compute
- OpenRouter vs Standard Compute
- OpenCode Go vs Standard Compute
- ClinePass vs Standard Compute
- Vercel AI Gateway vs Standard Compute
- Kilo Gateway vs Standard Compute
- DevPass vs Standard Compute
- Requesty vs Standard Compute
- LiteLLM vs Standard Compute
- Portkey vs Standard Compute
- Cloudflare AI Gateway vs Standard Compute
For production and governance
For companies
- Enterprise alternative to OpenRouterCost-optimized routing, a Europe rule, opt-in zero data retention, Entra SSO, per-key budgets and audit log.
- Enterprise alternative to LiteLLMNo Postgres or Redis to run, Entra SSO for everyone, a Europe rule and per-key budgets.
- Enterprise alternative to PortkeyModels and gateway in one contract, a Europe rule, Entra SSO and per-key budgets.
- Enterprise alternative to Kong AI GatewayNo per-model fee and no plugins to license, models included, a Europe rule, Entra SSO and budgets.
- Enterprise alternative to Azure OpenAIGPT, Claude, Gemini and open models in one API, no per-region TPM quota requests, Entra SSO and hard budgets.
- Enterprise alternative to AWS BedrockClaude, GPT, Gemini and open models in one API, no per-Region TPM quota, hard budgets per key and Entra SSO.
- Enterprise alternative to TrueFoundryModels included in one contract, a Europe rule, Entra SSO, per-key budgets and nothing to host.
- Enterprise alternative to Cloudflare AI GatewayA Europe rule per company, team or key, metadata-only request records, Entra SSO and SCIM.
- Enterprise alternative to Vercel AI GatewayA Vercel AI Gateway alternative: a Europe rule per company, team or key, opt-in zero data retention with no per-request fee, hard budgets and Entra SSO.
- Enterprise alternative to nexos.aiA Europe rule, opt-in zero data retention, Entra SSO with SCIM and per-key budgets.
- Enterprise alternative to LangdockOne API, a flat plan per person or app, a Europe rule and Entra SSO.
- Enterprise alternative to RequestyNo 5% markup, a Europe rule that covers inference, Entra SSO, per-key budgets and an audit log, self-serve.
- Enterprise alternative to EUrouterNo percentage markup on model spend, a Europe rule, Microsoft Entra SSO, teams, per-key budgets and audit log.
- Enterprise alternative to RegoloClaude, GPT and Gemini next to open models, a Europe rule, Microsoft Entra SSO, teams and per-key budgets.
- Enterprise alternative to HostYourAIClaude, GPT and Gemini next to open models, a Europe rule, Entra SSO and budgets that never overrun.
- Enterprise alternative to Prem AIPrem AI alternatives for companies: Standard Compute adds Claude, GPT and Gemini to open models on one API, with a Europe rule, Entra SSO and key budgets.
- Enterprise alternative to NodionClaude, GPT and Gemini next to European open models, Claude Code support, a Europe rule and Entra SSO.
- Enterprise alternative to KosmoyTeams, budgets, allow-lists, Entra SSO and an audit log without running a gateway in your Kubernetes.
- Enterprise alternative to OpperNo fee on credit purchases, routing that picks the model and provider, a Europe rule, Entra SSO and per-key budgets.
- Enterprise alternative to Orq.aiRouting to the cheapest capable model, a Europe rule on inference, Entra SSO and per-key budgets.
- Enterprise alternative to Maxim AI BifrostA managed alternative to Maxim AI's Bifrost gateway: nothing to host, frontier and open models on one account, a Europe rule, Entra SSO and budgets.
- Enterprise alternative to Mistral AIClaude, GPT and Gemini next to Mistral and open models, a Europe rule, Entra SSO, teams and per-key budgets.
Privacy
For companies: European clouds and company-wide AI
- Enterprise alternative to IONOS AI Model Hub
- Enterprise alternative to STACKIT AI Model Serving
- Enterprise alternative to Aleph Alpha
- Enterprise alternative to OVHcloud AI Endpoints
- Enterprise alternative to Scaleway Generative APIs
- Enterprise alternative to Infomaniak AI Services
- Enterprise alternative to Microsoft 365 Copilot
- Enterprise alternative to Cortecs