We use cookies to enhance your experience, provide personalised content, and analyse our traffic; see our Cookie Policy.
anthropic/claude-haiku-4-5-20251001Fast replies when the user is waiting
Anthropic describes Claude Haiku 4.5 as its fastest model with near-frontier intelligence. It offers a 200K-token context, up to 64K of output, tool use and vision over images and PDFs.
According to Anthropic, it reaches coding performance close to Sonnet 4 at more than twice the speed. It suits the turns a human waits on: a support bot, a form parser, message routing, tagging, a first-pass triage before a bigger model.
On Tchavi, a WhatsApp or web bot plugs in by changing your OpenAI SDK's base URL. You top up with mobile money, no card and no KYC, and your credits never expire.
Built for real-time replies, per Anthropic
Enough for a long conversation or a document of dozens of pages
Reads images and PDFs in the same request as your text
Function calling, with forced tool_choice accepted
Accepts temperature and top_p, which recent Claude models reject
A customer writes on WhatsApp: "My order is 40 minutes late, what are you doing about it?". Reply in two short, polite sentences, offering a concrete fix.
Sort this message into exactly one category from [payment, delivery, refund, other] and answer with the category only: "I paid with MTN MoMo but received nothing".
Translate to French for a shop page, warm tone: "Welcome to Mama Nana, your favourite wax prints delivered across Cotonou in 24 hours."
Haiku 4.5 when a human is waiting and the task is simple: support, triage, extraction. Sonnet 5.5, Anthropic's current Sonnet, when the answer needs deliberation, multi-step agents or a 1M context.
Haiku 4.5 to stay on Claude with a 200K context and forced tool_choice. Gemini 3.5 Flash if you need 1M tokens and native JSON mode; Google lists it among its legacy Flash models.
Yes. Anthropic only removed temperature and top_p from Opus 4.7 onward, and forced tool_choice is still accepted on Haiku 4.5, unlike Sonnet 5.5.
200K tokens in and up to 64K out, with images and PDFs.
With the OpenAI SDK, change the base URL, use your sk-tch-… key and model: "claude-haiku-4-5-20251001" (the full dated ID) on /v1/chat/completions.
Yes. Top up your Tchavi account with Orange Money, MTN MoMo, M-Pesa or another mobile money operator. No bank card is needed. Credits land as soon as the payment is confirmed and are deducted as you use them.
Yes. Keep your current OpenAI SDK, just change the base URL to Tchavi's (shown in the API tab) and use your Tchavi key. The rest of your code stays the same.
import Tchavi from '@tchavi/sdk';
const client = new Tchavi({ apiKey: 'API_KEY' });
const response = await client.chat.completions.create({
model: 'claude-haiku-4-5-20251001',
messages: [
{
"role": "user",
"content": "Hello!"
}
],
});
console.log(response.choices[0].message.content);
console.log('Credits used:', response.tchavi.credits_used);Every parameter this model accepts on the API. Check the SDK for full typings.
modelRequiredID of the chat model (e.g. `claude-haiku-4-5-20251001`).
messagesRequiredConversation messages. Each item has a `role` (system/user/assistant) and a `content` that's either a plain string or an array of content blocks (`text`, `image_url`, `document_url`).
temperature1Sampling temperature between 0 and 2. Higher values = more random.
max_tokensMaximum number of tokens to generate in the response.
streamfalseIf true, partial message deltas are streamed as Server-Sent Events.
toolsFunction-calling toolbox. Pass an array of `{ type: "function", function: { name, description, parameters } }`. Returned tool calls land in `choices[].message.tool_calls`. Only models with `supportsTools: true` accept this.
tool_choiceHow the model picks among tools: `"auto"` (default), `"none"` to force a plain text reply, or `{ type: "function", function: { name } }` to force a specific tool. Only with `supportsTools: true`.
top_p1Nucleus sampling: only tokens in the top `top_p` probability mass are considered.
stopUp to 4 sequences where the API will stop generating further tokens.
frequency_penalty0Between -2.0 and 2.0. Positive values penalize tokens that have appeared frequently.
presence_penalty0Between -2.0 and 2.0. Positive values encourage the model to talk about new topics.
messages[].content[].image_urlAttach an image on a user message. Pass a content block `{ "type": "image_url", "image_url": { "url": ... } }`. PNG / JPEG / WebP, up to 4 per request. Only models with `supportsImageInput: true` accept this block (e.g. `gpt-4o`, `claude-opus-4-7`).
messages[].content[].document_urlAttach a PDF on a user message. Pass a content block `{ "type": "document_url", "document_url": { "url": ... } }`. Up to 5 PDFs per request. Only models with `supportsDocuments: true` accept this block (e.g. `claude-sonnet-4-6`, `gpt-4.1`).
effort"high"Reasoning depth on Anthropic adaptive-thinking models (Opus 4.7+ / Sonnet 4.6+). Values `"low"`, `"medium"`, `"high"`. Higher = more deliberate, slower. Defaults to `"high"` on Opus 4.8.
Estimated over 30 days, for a typical message: ~500 input + ~150 output tokens, billed to the thousandth of a credit.