Gemma 4 vs Mercury 2.5 Preview
Google · US | Inception Labs · US · Updated June 2026
Quick verdict
Pick Gemma 4 for self-hosted, data-private deployment or running locally or on edge devices. Pick Mercury 2.5 Preview for very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation or coding accuracy (95.7%, 91st percentile among cost-optimized models). Choose Gemma 4 if you need self-hosting or data privacy; Mercury 2.5 Preview if you want a managed API.
Gemma 4 (Google) and Mercury 2.5 Preview (Inception Labs) are two of the models people most often weigh against each other in 2026. Gemma 4 is google's open-weight family: Apache 2.0 licensed, multimodal, and sized from edge devices up, for private self-hosting. Mercury 2.5 Preview is inception Labs' August 31, 2026 diffusion-based language model preview, refining tokens in parallel for roughly 10x the throughput of comparable autoregressive models at cost-optimized-tier quality. They diverge most on price, context window and open vs. closed weights — each quantified below from the models' real specs.
Key differences at a glance
- ▸Cost model: Gemma 4 ships open weights you can self-host (hardware cost only, no per-token fee), while Mercury 2.5 Preview is API-metered at $0.04/$0.15 per 1M tokens. Your choice depends on whether you want zero marginal cost at the price of running infrastructure.
- ▸Context window: 256K vs 260K tokens — within a few percent of each other, so treat this as a tie and test on your own long inputs, since usable recall varies by model.
- ▸Recency: Mercury 2.5 Preview is the newer model by about 5 months (released August 31, 2026), usually meaning fresher training data and capabilities.
Side-by-side specs
| Spec | Gemma 4 | Mercury 2.5 Preview |
|---|---|---|
| Provider | Google (US) | Inception Labs (US) |
| Released | April 2, 2026 | August 31, 2026 |
| Context window | 256K (~384 pages) | 260K tokens (~390 pages) |
| Price (in/out) | Open weight (self-host / free) | $0.04/$0.15 per 1M tokens |
| Open weight? | Yes — self-hostable | No — API only |
| Modalities | text, image, code | text |
| SWE-Bench Verified | Not published | Not published |
| MRCR v2 @ 1M | Not published | Not published |
Who wins what
Self-hosted, data-private deployment
Gemma 4
Open weights make this possible at all — Mercury 2.5 Preview is API-only, so it cannot leave the vendor's servers.
Running locally or on edge devices
Gemma 4
Google's open-weight family: Apache 2.0 licensed, multimodal, and sized from edge devices up, for private self-hosting — and its weights are open while Mercury 2.5 Preview is API-only.
Fine-tuning on your own data
Gemma 4
Gemma 4 lists fine-tuning on your own data among its strengths; Mercury 2.5 Preview does not.
Very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation
Mercury 2.5 Preview
Inception Labs' August 31, 2026 diffusion-based language model preview, refining tokens in parallel for roughly 10x the throughput of comparable autoregressive models at cost-optimized-tier quality — and it is the newer of the two.
Coding accuracy (95.7%, 91st percentile among cost-optimized models)
Mercury 2.5 Preview
Gemma 4 is comparatively weak here — trails frontier closed models on the hardest tasks
Mathematics accuracy (97.0%, 97th percentile)
Mercury 2.5 Preview
Mercury 2.5 Preview lists mathematics accuracy (97.0%, 97th percentile) among its strengths; Gemma 4 does not.
Lowest cost at scale
Gemma 4
Its weights are open, so at volume you pay for your own hardware instead of Mercury 2.5 Preview's $0.04/$0.15 per 1M tokens.
Which should you pick?
A cost-sensitive startup shipping high volume
→ Gemma 4
At Open weight (self-host / free) it undercuts Mercury 2.5 Preview, and on millions of tokens that margin decides the monthly bill.
Someone analysing very long documents or codebases
→ Mercury 2.5 Preview
Larger 260K tokens window fits more in one prompt.
A team with data-privacy or self-hosting needs
→ Gemma 4
Open weights let you run it on your own hardware; Mercury 2.5 Preview is API-only.
Anyone whose priority is self-hosted, data-private deployment
→ Gemma 4
It is specifically built for that.
Anyone whose priority is very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation
→ Mercury 2.5 Preview
That is its strongest area.
Gemma 4: where it fits
Google's open-weight family: Apache 2.0 licensed, multimodal, and sized from edge devices up, for private self-hosting. Released April 2, 2026 by Google, it is built for self-hosted, data-private deployment, running locally or on edge devices, fine-tuning on your own data, and multimodal tasks over a 256K context.
Its trade-offs are real: trails frontier closed models on the hardest tasks, and needs your own hardware to run. As an open-weight model, its running cost is your own hardware rather than a per-token fee.
Mercury 2.5 Preview: where it fits
Inception Labs' August 31, 2026 diffusion-based language model preview, refining tokens in parallel for roughly 10x the throughput of comparable autoregressive models at cost-optimized-tier quality. Released August 31, 2026 by Inception Labs, it is built for very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation, coding accuracy (95.7%, 91st percentile among cost-optimized models), mathematics accuracy (97.0%, 97th percentile), and tunable reasoning levels with parallel tool calls and schema-aligned JSON output.
Its trade-offs: 260K context window is far shorter than frontier 1M-token models, no vision or audio modalities, positioned only against cost-optimized models (GPT-5.6 Luna Low, Gemini 3.5 Flash-Lite), not frontier-class, and current $0.04/$0.15 pricing includes a limited-time launch promotion on OpenRouter (list price is $0.20/$0.75, may rise after Sept 8, 2026). At $0.04 in / $0.15 out per million tokens, it sits in the budget price band.
The bottom line for this matchup
The defining split here is open vs. closed. Gemma 4 gives you weights you control — self-host it, fine-tune it, keep data in-house, pay only for hardware. Mercury 2.5 Preview gives you a managed, always-updated API with no infrastructure to run. Teams with GPUs, privacy requirements, or huge volume often favour the open model; teams that want zero ops and the latest capabilities favour the closed one. Capability is close enough that this operational question, not the benchmark, usually decides it.
Want both Gemma 4 and Mercury 2.5 Preview without two subscriptions? LumiChats gives you these plus 40+ models under one ₹69/day pass (about $1/day) — draft with one, cross-check with the other.
See pricingFrequently asked questions
Is Gemma 4 or Mercury 2.5 Preview better for coding?
Public SWE-Bench figures are not available for either model, so the honest test is your own repository — run an identical real bug through both. By design, Gemma 4 leans toward self-hosted, data-private deployment while Mercury 2.5 Preview leans toward very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation, and that positioning usually predicts which feels better on your codebase.
Which is cheaper, Gemma 4 or Mercury 2.5 Preview?
Gemma 4 is open-weight, so self-hosting means no per-token fee (you pay for hardware instead), while Mercury 2.5 Preview is API-metered at $0.04/$0.15 per 1M tokens. For most teams without GPUs, the API model is cheaper to start; at very high volume, self-hosting can win.
Which has the bigger context window?
Effectively neither — 256K vs 260K tokens is a difference of a few percent. Remember advertised ≠ usable: recall typically degrades before the ceiling.
Can I use both Gemma 4 and Mercury 2.5 Preview together?
Yes — a multi-model platform like LumiChats gives you Gemma 4, Mercury 2.5 Preview and 40+ others under one ₹69/day pass (about $1/day), so you can draft with one and cross-check with the other instead of buying two subscriptions.
Which is newer, Gemma 4 or Mercury 2.5 Preview?
Mercury 2.5 Preview — released August 31, 2026, about 5 months after Gemma 4.
Related comparisons
Specifications and benchmarks reflect publicly reported figures as of June 2026 and may change as providers release updates. Always verify on your own workload.