Gemini 4 Argon vs Gemma 4 31B
Gemini 4 Argon scores higher on the Models at Work leaderboard (85 vs 58 out of 100). On Artificial Analysis' independent Intelligence Index, Gemini 4 Argon leads 53 to 15.
| Gemini 4 Argon · Google | Gemma 4 31B · Google | |
|---|---|---|
| Leaderboard rank | #3 | #22 |
| Score (0–100) | 85 | 58 |
| Tier | Frontier | Open weights |
| Price per 1M tokens | $1.99 blended (estimate) | Not verified |
| Context window | 1M | 256K |
| Open weights | No | Yes |
| Artificial Analysis Intelligence Index | 53 | 15 |
| Arena text score | 1525 | — |
| Arena Text score | — | 1452 (rank 75) |
| Best for | Chat and assistant products, Google Cloud shops, Price-sensitive frontier work | On-device and workstation inference, OCR and document vision, Deterministic pipeline/automation steps |
Choose Gemini 4 Argon if…
- you need chat and assistant products
- you need google cloud shops
- you need price-sensitive frontier work
Wins the popularity contest: first on Arena, mid-pack on the index, and priced like a workhorse.
Caveats: Google's public pricing page did not list Argon when read; the blended figure is Artificial Analysis' estimate.
Choose Gemma 4 31B if…
- you need on-device and workstation inference
- you need ocr and document vision
- you need deterministic pipeline/automation steps
The dense Apache-2.0 model local-first teams call 'the new baseline'. Excellent at rule-following, OCR and automation; needs ~48GB for full 256K context and is not a frontier reasoner. Free to run, no API price to speak of.
Caveats: Dense 31B: expect ~12-16 tok/s on Apple Silicon at Q4-Q6, and 70GB RAM at full context. Several HN users say Qwen 3.6/3.8 handles long context and agentic tasks a little better; Gemma wins on instruction discipline. AA 15 on the current index; this is a local model, not a hosted-API competitor.
What practitioners say
Gemini 4 Argon: The announcement thread ran to 1,190 comments; the independent analysis thread was smaller and more measured.
Gemma 4 31B: HN's local-LLM crowd treats it as the reference point: 'the new baseline for local models' (soganess, 70GB peak on an M5 Max at 256K context); 'particularly good at pipeline/automation tasks' and better than Qwen even at 100B+ for rule-following; 'very good at OCR'. 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 ~11-16 tok/s on Macs feels slow.
Questions people ask
Which is better, Gemini 4 Argon or Gemma 4 31B?
Gemini 4 Argon scores higher on the Models at Work leaderboard (85 vs 58 out of 100). On Artificial Analysis' independent Intelligence Index, Gemini 4 Argon leads 53 to 15.
When should I choose Gemini 4 Argon over Gemma 4 31B?
Choose Gemini 4 Argon for chat and assistant products, google cloud shops, price-sensitive frontier work. Wins the popularity contest: first on Arena, mid-pack on the index, and priced like a workhorse.
When should I choose Gemma 4 31B over Gemini 4 Argon?
Choose Gemma 4 31B for on-device and workstation inference, ocr and document vision, deterministic pipeline/automation steps. The dense Apache-2.0 model local-first teams call 'the new baseline'. Excellent at rule-following, OCR and automation; needs ~48GB for full 256K context and is not a frontier reasoner. Free to run, no API price to speak of.
Scores come from the Models at Work leaderboard: Score is 0 to 100 and computed, not typed: 55% independent evals (Artificial Analysis, Arena, Scale SEAL, Epoch, each scaled against its natural floor and the best score in this table, then averaged), 15% blended price on a fixed log scale ($0.05 per million is 100, $60 is 0), 15% practitioner sentiment from Signals and community threads, 15% operational fit (context window, open weights). A missing component drops out and the rest are reweighted. A model with no independent eval 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; blended figures assume three input tokens per output token.