Home / Learn

AI API knowledge hub

Pick the right model, speak the protocols, understand the bill — plain-English guides for the model API world.

1 · Which model should I pick?

There is no single “best model” in 2026 — there are best fits. The fastest way to choose is by workload type:

WorkloadPickWhy
Hardest reasoning
research, math, architecture
Claude Opus 5 / GLM-5.3Deepest chain-of-thought quality; Opus 5 carries a full 1M-token context window.
Daily coding
IDE agents, code review
Claude Sonnet 5 / DeepSeek V4Best speed-to-quality balance in tools like Claude Code and Codex; DeepSeek V4 is the open-weights value pick.
High-volume / bulk
classification, extraction, batch
Haiku 4.5 / DeepSeek V4 Flash / Qwen 3.8 FlashFraction of the cost with strong instruction following.
Multilingual workflows
CJK content, long documents, localisation
Qwen 3.8 / Kimi K3 / GLMStrongest CJK comprehension and generation; Kimi excels at very long context.
Vision / multimodalQwen 3-VL PlusImage understanding at a fraction of flagship vision pricing.
Codex / GPT ecosystemGPT-5.5 / 5.4Native fit for OpenAI-aligned agent frameworks.

Rule of thumb: start with the middle tier (Sonnet 5 / DeepSeek V4) for real work, escalate to Opus-class only when the task proves hard, and drop to Haiku/Flash-class for anything repetitive. Your bill shrinks ~10×, quality loss is often zero.

2 · One gateway, three protocols

Model APIs speak different wire formats. ThunderPeak exposes them all behind one key and one balance:

ProtocolEndpointUsed by
OpenAI Chat Completions/v1/chat/completionsOpenAI SDK, Cherry Studio, most tools
Anthropic Messages/v1/messagesClaude Code, Anthropic SDK, Cursor
Responses/v1/responsesCodex CLI & extension, newer OpenAI SDKs

Point an OpenAI-compatible tool at https://thunder-peak.com/v1, or an Anthropic tool at https://thunder-peak.com — the same key works on both. If a third-party client reports a path error, append /v1.

3 · How the bill is calculated

Every request meters five token categories — this is the same model the providers use, so our per-token prices map 1:1 to what you'd expect:

CategoryMeaningRelative cost
Input tokensYour prompt, system prompt, tool definitions
Output tokensThe generated answertypically 5× input
Cache writeFirst time a prompt prefix is cached1.25× input
Cache readReusing a cached prefix (repeat prompts)0.1× input — biggest lever
Thinking tokensHidden reasoning (Opus/Sonnet/thinking models)output-priced

Practical tips: keep your system prompt stable to maximise cache reads (10× cheaper than fresh input); streaming costs the same as non-streaming; and usage history in the dashboard breaks every request down to the token.

4 · Why route through a gateway at all?

⚡ One key for every model
No per-provider accounts, no per-provider billing dashboards. Claude, GPT, DeepSeek, Qwen, Kimi and GLM share one balance.
💳 No overseas card needed
Pay with USDT on TRC-20 / BSC / Solana. Top up $10 or $10,000 — same process.
🎯 Fixed discount on flagships
Claude & GPT at 66% of official rates; other major models at official rates — the discount is a policy, not a promo that expires.
🛡 Per-key control
Issue separate keys for each project with individual caps. Freeze or delete them from the console in one click.

5 · Glossary

TermPlain meaning
API keyYour secret credential; send it as Authorization: Bearer sk-tp-…
Base URLThe gateway address you paste into a tool, e.g. https://thunder-peak.com/v1
Context windowMax tokens the model can see per request (e.g. 1M ≈ a whole codebase)
SSE / streamingToken-by-token output; what makes chat and agent UIs feel instant
Tool callingThe model asking your app to run functions — the basis of agents
Reasoning / thinkingModels that "think" before answering; stronger but pricier per token
Rate limitRequests per minute allowed per key — configurable per key in ThunderPeak

Still choosing? Try them side by side.

Sign up, get $2.50 free, and compare models yourself.