MeituanWeights not listed·1000K context·— params

LongCat-2.5-Previewmeituan/longcat-2-5-preview

Name: LongCat-2.5-Preview API
Brand: Meituan
Availability: InStock

Meituan's September 25 preview adds image understanding and is offered through OpenAI- and Anthropic-compatible APIs. The published USD prices are a limited-time discount. We have not verified downloadable weights, a license or a discount end date for this exact version.

Cheapest blended:$0.53 / 1M tokenson Meituan LongCat · 1 provider listed

Official sources checked October 2, 2026. Evaluation suggestions below are not benchmark results.

Preview access and temporary pricing

The official September 25 change log announces this preview with image understanding. Use the exact API ID LongCat-2.5-Preview. Its pricing page labels the displayed USD rates as a limited-time discount without stating an end date; do not treat them as a fixed long-term commitment.

The platform is operated by Beijing Kuxun Interactive Technology and affiliates under Meituan. Account creation, API access and payment eligibility need separate verification. The availability of English documentation is not proof that every overseas account can complete signup or billing.

Endpoint details that affect integration

The official cURL example uses https://api.longcat.chat/openai/v1/chat/completions. This directory uses the matching /openai/v1 base in generated examples. The docs also show an SDK example without /v1; check the actual resolved request path if your SDK produces a 404. Anthropic-format requests use a separate /anthropic route.

Start with a short text-only request, then test tools and image input separately. That isolates request-format errors from model-quality failures. Check status codes, returned usage and unsupported parameters before porting a full agent workflow.

Evaluate the preview before replacing a stable route

For screenshot-assisted coding, keep a small set of cases where the screenshot changes the correct fix. Score whether the model uses the visible evidence, introduces unrelated changes or needs extra repair attempts. Add text-only controls to avoid crediting image support for tasks that did not need it.

A 7.5M-input/2.5M-output workload with no cache hits costs $5.25 at the displayed discount. This calculated scenario excludes retries and additional charges. Record real usage and success rates before choosing the route. No comparative benchmark or downloadable 2.5 checkpoint has been verified here.

Official source 1 →Official source 2 →Official source 3 →Official source 4 →

Pricing across providers

Sort by:

Provider	Input /1M	Output /1M	Blended /1M	Latency p50	Format	Freshness	Action
Meituan LongCat LongCat-2.5-Preview Limited-time discounted USD pay-as-you-go rate. No discount end date published; billing records are authoritative.	$0.3 Cached $0.006	$1.2	$0.525	—	OpenAI-compatible	Needs 24d ago	Try →

Affiliate disclosure: We may earn a commission from qualified signups. Pricing independence is enforced at the data layer — see our Editorial Independence Policy.

Works with

Point any of these clients at a hosting's base URL — they all speak at least one of this model's endpoint protocols (OPENAI_COMPATIBLE).

Capabilities

reasoning
coding
tool_use
image_understanding

Code samples

Example using Meituan LongCat — the cheapest hosting for this model as of last verification. Swap base_url and model to use a different provider from the matrix above.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.longcat.chat/openai/v1",
)

response = client.chat.completions.create(
    model="LongCat-2.5-Preview",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

Technical specs

Context: 1000K
Max output: 128K
Parameters: —
Release: 2026-09-25
Training cutoff: —
License: —

Official documentation →

Similar models

Compare with

LongCat-2.5-Preview vs DeepSeek V4.1 Flash
Comparison planned — not yet published
LongCat-2.5-Preview vs GLM 5.3 Flash
Comparison planned — not yet published
LongCat-2.5-Preview vs Kimi K2.5
Comparison planned — not yet published

Frequently asked

How much does LongCat-2.5-Preview cost?+

The cheapest public hosting is $0.53 per 1M blended tokens on Meituan LongCat. 1 total providers are listed above with per-input / per-output / cached pricing.

Is LongCat-2.5-Preview open-source?+

Downloadable weights are not listed in this directory for LongCat-2.5-Preview. See the official documentation and availability notes for API, subscription or planned weight-release details.

Is LongCat-2.5-Preview OpenAI-compatible?+

At least one listed hosting exposes an OpenAI-compatible API, so you can point an existing openai SDK client at the Provider's base_url and use the Provider's model name. See the Code Samples above for a copy-pasteable example.

What's the maximum context window for LongCat-2.5-Preview?+

The model supports up to 1,000,000 tokens of context (input + output). Some hosted versions may impose a smaller limit — check the "Context" column in the pricing matrix for each provider.