@cf/mistralai/mistral-small-3.1-24b-instruct vs GLM 5.2

Practical read: @cf/mistralai/mistral-small-3.1-24b-instruct costs about half as much. GLM 5.2 reads eight times more context. One API key runs both.

No card required. Start with 1 million free tokens.

Side by side:

@cf/mistralai/mistral-small-3.1-24b-instruct compared with Z.ai: GLM 5.2
Attribute	@cf/mistralai/mistral-small-3.1-24b-instruct `mistralai/mistral-small-3.1-24b-instruct`	GLM 5.2 `z-ai/glm-5.2`
Price, 1M in + out	$0.96 Cloudflare Workers AI, cheapest route	$2.30 Engy, cheapest route
Best for	Low cost	Long documents
Input / output, 1M	$0.37 / $0.59	$0.72 / $1.58
Cached input	Not published	$0.131875/1M to $0.773315/1M
Context window	128K 128,000 tokens	1.05M 1,048,576 tokens
First token, p50	Measuring	0.66 s via FriendliAI
Throughput	Measuring	265 tok/s via Telnyx
AI IQ	Unrated	IQ 118 #53 on AI IQ
Provider routes	1	31 with failover
Privacy	Varies by provider	Confidential E2EE and zero retention route
Recent route uptime range	100.00%	0.00% to 100.00%
Publisher	Mistral	Z.AI
Modes	chat	chat
More detail	Overview	Overview Pricing Providers Benchmarks

Price adds one million input and one million output tokens on the cheapest route. Prices and speed change: see the monthly reports and run a small eval before you choose a default.

Test both with your agent

Paste this into Claude Code, Cursor or any coding agent.

Compare @cf/mistralai/mistral-small-3.1-24b-instruct and GLM 5.2 in this project through TrustedRouter.

1. Read https://trustedrouter.com/docs/llms.txt
2. Point the OpenAI-compatible client at
   https://api.trustedrouter.com/v1 and read the key from
   TRUSTEDROUTER_API_KEY.
3. Run our eval prompts on mistralai/mistral-small-3.1-24b-instruct, then on z-ai/glm-5.2.
4. Show me cost, latency and quality side by side before changing the default.

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.trustedrouter.com/v1",
    api_key=os.environ["TRUSTEDROUTER_API_KEY"],
)
for model in ("mistralai/mistral-small-3.1-24b-instruct", "z-ai/glm-5.2"):
    response = client.chat.completions.create(
        model=model,
        messages=[{"role": "user", "content": "What is the capital of France?"}],
    )
    print(model, response.choices[0].message.content)

for model in mistralai/mistral-small-3.1-24b-instruct z-ai/glm-5.2; do
  curl https://api.trustedrouter.com/v1/chat/completions \
    -H "Authorization: Bearer $TRUSTEDROUTER_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{\"model\": \"$model\", \"messages\": [{\"role\": \"user\", \"content\": \"What is the capital of France?\"}]}"
done

Your agent will need an API key.

Questions

Which should I use, @cf/mistralai/mistral-small-3.1-24b-instruct or Z.ai: GLM 5.2?

@cf/mistralai/mistral-small-3.1-24b-instruct has the lower published input-plus-output rate. Z.ai: GLM 5.2 has more Credits provider routes and the larger context window. There is not enough recent probe data to compare speed.

Is @cf/mistralai/mistral-small-3.1-24b-instruct or Z.ai: GLM 5.2 cheaper?

The current cheapest TrustedRouter route is $0.95583/1M for @cf/mistralai/mistral-small-3.1-24b-instruct and $2.2999/1M for Z.ai: GLM 5.2. The comparison uses current catalog prices and updates as provider pricing changes.

Is @cf/mistralai/mistral-small-3.1-24b-instruct or Z.ai: GLM 5.2 faster?

Current measured p50 time to first token is not enough data for @cf/mistralai/mistral-small-3.1-24b-instruct and 659 ms for Z.ai: GLM 5.2. These are routed probe measurements, not vendor-advertised speeds, and update as new samples arrive.

Can I test @cf/mistralai/mistral-small-3.1-24b-instruct and Z.ai: GLM 5.2 with the same API?

Yes. Use the same OpenAI-compatible TrustedRouter base URL and API key, then change only the model id between mistralai/mistral-small-3.1-24b-instruct and z-ai/glm-5.2. This makes side-by-side evals possible without maintaining two provider integrations.

More comparisons

All comparisons

Try @cf/mistralai/mistral-small-3.1-24b-instruct and GLM 5.2 with one key.

TrustedRouter is useful when you know the model you want, but still need provider rollover, budget limits, usage records, and a prompt path you can verify.

No card required. Start with 1 million free tokens.