ERNIE 5.0 vs Mercury 2.5 Preview

Baidu · China  |  Inception Labs · US · Updated June 2026

Quick verdict

Pick ERNIE 5.0 for baidu's flagship omni-modal model — text, image and video understanding or particularly strong on chinese-language reasoning tasks. 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). On a tight budget at scale, Mercury 2.5 Preview is the value pick.

ERNIE 5.0 (Baidu, China) and Mercury 2.5 Preview (Inception Labs, US) line up two different AI ecosystems against each other — a comparison that is as much about cost philosophy and openness as raw capability. ERNIE 5.0 is baidu's flagship omni-modal ERNIE model — strong Chinese-language and multimodal reasoning at low cost, though it trails the Western frontier on independent benchmarks. 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 and context window — each quantified below from the models' real specs.

Key differences at a glance

Side-by-side specs

SpecERNIE 5.0Mercury 2.5 Preview
ProviderBaidu (China) Inception Labs (US)
ReleasedJanuary 22, 2026 August 31, 2026
Context window128K (~192 pages) 260K tokens (~390 pages)
Price (in/out)$0.6/$2.1 per 1M tokens $0.04/$0.15 per 1M tokens
Open weight?No — API only No — API only
Modalitiestext, image, video, code text
SWE-Bench VerifiedNot published Not published
MRCR v2 @ 1MNot published Not published

Who wins what

Baidu's flagship omni-modal model — text, image and video understanding

ERNIE 5.0

Mercury 2.5 Preview is comparatively weak here — 260K context window is far shorter than frontier 1M-token models

Particularly strong on Chinese-language reasoning tasks

ERNIE 5.0

ERNIE 5.0 lists particularly strong on Chinese-language reasoning tasks among its strengths; Mercury 2.5 Preview does not.

Competitive API pricing (around $0.60/$2.10 per million tokens)

ERNIE 5.0

Mercury 2.5 Preview is comparatively weak here — 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)

Very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation

Mercury 2.5 Preview

Its 260K tokens window holds about 2× more than ERNIE 5.0's 128K in a single prompt.

Coding accuracy (95.7%, 91st percentile among cost-optimized models)

Mercury 2.5 Preview

At $0.04/$0.15 per 1M tokens it undercuts ERNIE 5.0 ($0.6/$2.1 per 1M tokens), and that gap compounds at volume.

Mathematics accuracy (97.0%, 97th percentile)

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 runs cheaper at $0.04/$0.15 per 1M tokens.

Lowest cost at scale

Mercury 2.5 Preview

At $0.04/$0.15 per 1M tokens, it is the cheaper of the two — the gap dominates the bill on high-volume workloads.

Largest single-prompt input

Mercury 2.5 Preview

Its 260K tokens window is about 2× larger than ERNIE 5.0's 128K, fitting roughly 390 pages in one prompt.

Which should you pick?

A cost-sensitive startup shipping high volume

Mercury 2.5 Preview

At $0.04/$0.15 per 1M tokens it undercuts ERNIE 5.0, 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.

Anyone whose priority is baidu's flagship omni-modal model — text, image and video understanding

ERNIE 5.0

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.

An enterprise with regional data-residency rules

Mercury 2.5 Preview or ERNIE 5.0

Origin (China vs US) affects where data is processed and which compliance regime applies — check the provider's terms for your region.

ERNIE 5.0: where it fits

Baidu's flagship omni-modal ERNIE model — strong Chinese-language and multimodal reasoning at low cost, though it trails the Western frontier on independent benchmarks. Released January 22, 2026 by Baidu, it is built for baidu's flagship omni-modal model — text, image and video understanding, particularly strong on Chinese-language reasoning tasks, competitive API pricing (around $0.60/$2.10 per million tokens), and backed by a major lab with deep China-market integration.

Its trade-offs are real: trails the Western frontier on aggregate independent tests (AA Intelligence Index ~22 for the tracked Thinking Preview), parameter and architecture details are vendor-stated and opaque, closed weights on a China-hosted API, and 128K context is smaller than 1M-token rivals. At $0.6 in / $2.1 out per million tokens, it sits in the budget price band.

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

This is less "which is smarter" and more "which ecosystem fits." ERNIE 5.0 (China) and Mercury 2.5 Preview (US) differ on pricing philosophy, data-residency, and tooling as much as on raw scores. Mercury 2.5 Preview is the cheaper option, which matters at volume. The pragmatic move is to run one real task through both and judge the outputs against your own constraints — including where your data is allowed to be processed.

Want both ERNIE 5.0 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 pricing

Frequently asked questions

Is ERNIE 5.0 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, ERNIE 5.0 leans toward baidu's flagship omni-modal model — text, image and video understanding 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, ERNIE 5.0 or Mercury 2.5 Preview?

Mercury 2.5 Preview is cheaper — $0.6/$2.1 per 1M tokens vs $0.04/$0.15 per 1M tokens, roughly 15× apart on input.

Which has the bigger context window?

Mercury 2.5 Preview — 260K tokens vs 128K, about 2× larger. Useful only if the model actually reasons over the full window, which not all do.

Can I use both ERNIE 5.0 and Mercury 2.5 Preview together?

Yes — a multi-model platform like LumiChats gives you ERNIE 5.0, 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, ERNIE 5.0 or Mercury 2.5 Preview?

Mercury 2.5 Preview — released August 31, 2026, about 7 months after ERNIE 5.0.

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.