Gemini 4 Carbon vs Gemma 4 31B
Gemini 4 Carbon scores higher on the Models at Work leaderboard (75 vs 54 out of 100). On independent tests alone, Gemini 4 Carbon is ahead (92 vs 37). Gemma 4 31B has no list price we could verify, so the two cannot be compared on cost. On the Artificial Analysis Intelligence Index, one of the independent tests, Gemini 4 Carbon leads 60 to 4.
| Gemini 4 Carbon · Google | Gemma 4 31B · Google | |
|---|---|---|
| Leaderboard rank | #1 | #25 |
| Score (0–100) | 75 | 54 |
| Kind of work | Hardest problems | Run it yourself |
| Price per 1M tokens | $2 to read · $10 to write | Not verified |
| Reads at once | 1M tokens (about 750,000 words) | 256K tokens (about 192,000 words) |
| Where you can get it | Gemini Enterprise Agent Platform (Google Cloud), Google AI Studio, Gemini for business, Gemini app (Google AI Ultra) | Amazon Bedrock, Gemini Enterprise Agent Platform (Google Cloud) |
| Can a company host it itself | No | Yes (open weights) |
| Artificial Analysis Intelligence Index | 60 | 4 |
| Arena Text score | Not measured | 1452 (rank 75) |
| Best for | Reading and comparing large sets of dense documents, Legal and finance analysis, Multi-step research | Reading scanned documents, Fixed, repeatable steps in a process, Keeping data on your own machines |
Cost of 1,000 jobs
| 1,000 of these | Gemini 4 Carbon | Gemma 4 31B |
|---|---|---|
| Support repliesthe customer message, the account notes and two help articles read; a 300-word reply written | $10 | No list price |
| Documents processeda two-page invoice or form read; the fields pulled out | $7 | No list price |
| Calls summariseda 40-minute call transcript read; a summary and next steps written | $26 | No list price |
| Contracts revieweda 40-page contract read against a checklist; a memo of issues written | $75 | No list price |
| Drafts writtena brief and one example read; a 600-word email or proposal section written | $11 | No list price |
Worked out from each vendor’s list price and a fixed size for each job. A product built on a model charges its own price on top. See what every model costs for the same jobs.
Choose Gemini 4 Carbon for
- Reading and comparing large sets of dense documents
- Legal and finance analysis
- Multi-step research
Second on Artificial Analysis's index (60 against the leader's 65) at a quarter of the leader's list price, but only until 30 November, when Google doubles it.
Caveats: The $2 in / $10 out per million tokens (the unit vendors bill by) is introductory; Google says $4 / $20 applies after 30 November 2026, which will lower its score here. Google says it uses more text per task than Gemini 4 Argon, so the bill can run higher at the same list price. One independent test so far. The rankings quoted in Google's launch post are relayed by Google and not counted. In the Gemini app it is for Google AI Ultra subscribers only, per Google; commenters on lower plans and one enterprise account report they cannot select it yet.
Choose Gemma 4 31B for
- Reading scanned documents
- Fixed, repeatable steps in a process
- Keeping data on your own machines
Google's small open model for fixed, rule-following steps: free to use commercially and able to run on one strong workstation, but 4 on Artificial Analysis's index rules out reasoning.
Caveats: Slow on a desktop: users report 12 to 16 tokens (word-pieces) a second on Apple machines, and about 70GB of memory to read its longest documents. Users say the rival Qwen models handle long documents and multi-step tasks a little better; Gemma is better at sticking to instructions. Amazon Bedrock lists it at $0.14 in / $0.40 out per million tokens in US regions if you would sooner rent than run it.
What people running them say
Gemini 4 Carbon: On the Hacker News (a developer forum) thread about Google's launch, several commenters on the Pro and Plus plans say Carbon shows in the app as Ultra-only, one says their company's enterprise account cannot use it yet, and another says their workplace already finds Argon too expensive. None of this is verified, and it is too little to score.
Gemma 4 31B: Hacker News (a developer forum) users who run models on their own computers treat it as the reference point: 'the new baseline for local models' (soganess, who saw it use 70GB of memory on an Apple M5 Max); 'particularly good at pipeline/automation tasks' and better at following rules than far larger Qwen models; 'very good at OCR' (reading text from images). One user had it catch a bug Opus 4.7 missed. Counterpoints: Qwen 3.6 'handles context a little better', peers 'predominantly run Qwen', and about 11 to 16 tokens (word-pieces) a second on Macs feels slow.
Questions people ask
Which is better, Gemini 4 Carbon or Gemma 4 31B?
Gemini 4 Carbon scores higher on the Models at Work leaderboard (75 vs 54 out of 100). On independent tests alone, Gemini 4 Carbon is ahead (92 vs 37). Gemma 4 31B has no list price we could verify, so the two cannot be compared on cost. On the Artificial Analysis Intelligence Index, one of the independent tests, Gemini 4 Carbon leads 60 to 4.
When should I choose Gemini 4 Carbon over Gemma 4 31B?
Choose Gemini 4 Carbon for reading and comparing large sets of dense documents, legal and finance analysis, multi-step research. Second on Artificial Analysis's index (60 against the leader's 65) at a quarter of the leader's list price, but only until 30 November, when Google doubles it.
When should I choose Gemma 4 31B over Gemini 4 Carbon?
Choose Gemma 4 31B for reading scanned documents, fixed, repeatable steps in a process, keeping data on your own machines. Google's small open model for fixed, rule-following steps: free to use commercially and able to run on one strong workstation, but 4 on Artificial Analysis's index rules out reasoning.
Scores come from the Models at Work leaderboard: Score is 0 to 100 and computed, not typed. 40% is independent testing (Artificial Analysis, Arena, Scale, Epoch, Vals and others; each test is scaled against the best score on this board, and tests of real work, such as finished freelance projects or professional tasks, count double the exams). 30% is price: the vendor's list price for a typical mix of reading and writing, on a fixed scale from $0.05 to $60 per million tokens, so a model that costs a fifth as much gains real ground. 15% is what people report after running it, from Signals and linked threads. 15% is fit: how much text it can take in at once and where a business can get it (the products and clouds that carry it), with a small credit for open weights. A missing component drops out and the rest are reweighted. A model with no independent test yet is provisional and ranks below every measured one. Within three points is a tie. Vendor-published figures are labelled and never counted. Prices are vendor list prices where published; a typical mix assumes three parts reading to one part writing.