We use cookies. Cookies and browser storage keep you signed in and remember your preferences. With your permission we also measure how the site is used. You can any time. See our cookie policy.

New here? See how it works A 1-minute tour of the playground, models and your API key.

Command A+ vs GLM 5.3 FlashX: API price, context & features · api14

Cohere: Command A+ vs Z.ai: GLM 5.3 FlashX

A side-by-side comparison of Cohere: Command A+ ($1.05 / 1M tokens) and Z.ai: GLM 5.3 FlashX ($1.30 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.

Specifications

	Cohere: Command A+	Z.ai: GLM 5.3 FlashX
Provider	Cohere	Z.ai
Input	$1.05 / 1M tokens	$1.30 / 1M tokens
Output	$5.25 / 1M tokens	$4.38 / 1M tokens
Cache read	$0.52 / 1M tokens	$0.26 / 1M tokens
Context window	192,000	1,048,576
Max output	64,000	131,072
Input	text, image	text, image, video
Output	text	text
Tool calling
Structured output
Reasoning control
Added	Sep 22, 2026	Sep 18, 2026

Key differences

Cohere: Command A+ is 19% cheaper on input tokens.
Z.ai: GLM 5.3 FlashX is 17% cheaper on output tokens.
Z.ai: GLM 5.3 FlashX reads 5.5× more context (1,048,576 tokens).
Only Cohere: Command A+ supports structured JSON output.
A request with 1,000 input and 500 output tokens costs $0.00367 with Cohere: Command A+ and $0.00348 with Z.ai: GLM 5.3 FlashX.

In depth

Cohere: Command A+ is a text generation model from Cohere, released on 22 September 2026. Z.ai: GLM 5.3 FlashX is a text generation model from Z.ai, released on 18 September 2026. Both are served on api14 as text category models, and both take non text input alongside text and return text output, so the choice between them comes down to context size, modality coverage and the parameter surface each one exposes.

The clearest split is context length. GLM 5.3 FlashX offers a context window of 1,048,576 tokens, while Command A+ offers 192,000 tokens, so GLM 5.3 FlashX is the larger context option by a wide margin. On input modalities, Command A+ accepts text and images, while GLM 5.3 FlashX accepts text, images and video, making it the broader option if your pipeline includes video frames or clips. Both output text only. On parameters, Command A+ supports structured outputs, which GLM 5.3 FlashX does not list, and Command A+ also exposes frequency_penalty, presence_penalty, seed and stop, which GLM 5.3 FlashX does not. GLM 5.3 FlashX in turn lists reasoning_effort and tool_choice, which Command A+ does not. Both support tools, reasoning, include_reasoning, response_format, temperature, top_k, top_p and max_tokens, so tool calling and reasoning are available on either side.

For practical selection, reach for GLM 5.3 FlashX when the job involves very large inputs, such as whole repositories, long document sets or extended conversation histories, or when you need to pass video as input. Its reasoning_effort parameter also gives you a direct dial for how much reasoning work the model does. Reach for Command A+ when you need structured outputs, since that is the one capability it lists and GLM 5.3 FlashX does not, or when you want finer control over sampling through frequency_penalty, presence_penalty, seed and stop. If your inputs fit comfortably inside 192,000 tokens and are text or images only, both models can serve the same request, and the decision then rests on whether structured output or the extra sampling controls matter to you.

Calling either model through api14 works the same way, because the API is OpenAI compatible. You point your client at the api14 base URL, send your request to the chat completions endpoint, and set the model field to the id you want. To use Cohere's model, set the model id to cohere/command-a-plus. To use Z.ai's model, set the model id to z-ai/glm-5.3-flashx. Everything else in the request body, your messages, tools, response_format and sampling parameters, stays as it is, subject to the parameters each model supports. Switching between the two is a one line change to the model id, so you can route different workloads to different models without rewriting your integration.

There are cases where the other model is still the better pick. If you need structured outputs, GLM 5.3 FlashX is not the right choice, because that capability is listed only for Command A+. If you need a context window beyond 192,000 tokens, or you need to send video, Command A+ is not the right choice, because those belong to GLM 5.3 FlashX. On price, Command A+ is the cheaper option for input and GLM 5.3 FlashX is the cheaper option for output, so the cost balance depends on your ratio of input to output tokens; check the live price table on this page for the current figures.

Switch with one string

Both models use the same endpoint and request shape. Change the model value from "cohere/command-a-plus" to "z-ai/glm-5.3-flashx" and prompts, streaming and tools keep working.

Cohere: Command A+ details & playground →Z.ai: GLM 5.3 FlashX details & playground →

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api14.com/api/v1",
  apiKey: process.env.API14_API_KEY,
});

const result = await client.chat.completions.create({
  "model": "z-ai/glm-5.3-flashx",
  "messages": [
    {
      "role": "user",
      "content": "Explain vector databases in two sentences."
    }
  ]
});
console.log(result.choices[0].message.content);

Cohere: Command A+ vs Z.ai: GLM 5.3 FlashX

Specifications

Key differences

In depth

Switch with one string

More comparisons