Gemma 4 31B vs grok-4.20-0309-reasoning

Practical read: Gemma 4 31B costs about a ninth as much and starts answering sooner. One API key runs both.

No card required. Start with 1 million free tokens.

Side by side:

Google: Gemma 4 31B compared with grok-4.20-0309-reasoning
Attribute	Gemma 4 31B `google/gemma-4-31b-it`	grok-4.20-0309-reasoning `x-ai/grok-4.20-reasoning`
Price, 1M in + out	$0.46 Weights & Biases, cheapest route	$3.96 Microsoft Azure AI Foundry, cheapest route
Best for	Low cost and speed
Input / output, 1M	$0.11 / $0.36	$1.32 / $2.64
Cached input	$0.060927/1M to $0.26375/1M	$0.211/1M to $0.422/1M
Context window	262K 262,144 tokens	Unlisted
First token, p50	0.78 s via Tinfoil	1.4 s via xAI Grok
Throughput	250 tok/s via FriendliAI	Measuring
AI IQ	IQ 101 #120 on AI IQ	Unrated
Provider routes	14 with failover	2 with failover
Privacy	Confidential E2EE and zero retention route	Zero retention route
Recent route uptime range	100.00%	100.00%
Publisher	Google AI Studio	xAI Grok
Modes	chat	chat
More detail	Overview Pricing Providers	Overview Pricing Providers

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 Gemma 4 31B and grok-4.20-0309-reasoning 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 google/gemma-4-31b-it, then on x-ai/grok-4.20-reasoning.
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 ("google/gemma-4-31b-it", "x-ai/grok-4.20-reasoning"):
    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 google/gemma-4-31b-it x-ai/grok-4.20-reasoning; 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, Google: Gemma 4 31B or grok-4.20-0309-reasoning?

Google: Gemma 4 31B has the lower published input-plus-output rate, more Credits provider routes and the lower measured p50 time to first token.

Is Google: Gemma 4 31B or grok-4.20-0309-reasoning cheaper?

The current cheapest TrustedRouter route is $0.4642/1M for Google: Gemma 4 31B and $3.95625/1M for grok-4.20-0309-reasoning. The comparison uses current catalog prices and updates as provider pricing changes.

Is Google: Gemma 4 31B or grok-4.20-0309-reasoning faster?

Current measured p50 time to first token is 781 ms for Google: Gemma 4 31B and 1356 ms for grok-4.20-0309-reasoning. These are routed probe measurements, not vendor-advertised speeds, and update as new samples arrive.

Can I test Google: Gemma 4 31B and grok-4.20-0309-reasoning with the same API?

Yes. Use the same OpenAI-compatible TrustedRouter base URL and API key, then change only the model id between google/gemma-4-31b-it and x-ai/grok-4.20-reasoning. This makes side-by-side evals possible without maintaining two provider integrations.

More comparisons

All comparisons

Try Gemma 4 31B and grok-4.20-0309-reasoning 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.