The State of Translation Automation 2026 · Languages
What it takes to automate translation from English into Azerbaijani against 10 requirements in three levels: the solutions that meet them, what they cost and how fast they are, from the 2026 benchmark of 57 AI models and MT engines.
A solution is a model plus its harness: prompts, glossaries and other context, pre- and post-processing. It meets a quality level when it meets every requirement of the level.
| Requirement | Description | Met when |
|---|---|---|
| Basic requirements for any company | ||
| General translation quality | Translation errors of every type, such as mistranslations, omissions and grammar errors, classified by type and severity following MQM, the industry standard. | No critical error, at most 3 major errors in the four test texts (3,034 words) and at most 20 errors per 1,000 words, weighted by severity (minor 1, major 5, critical 10). |
| Standard requirements adds the rules companies set for their content | ||
| Terminology | The approved translation of every glossary term, in any of its correct forms. | 100% of the glossary terms. We test it in the 16 languages with a glossary, so it does not apply in Azerbaijani. |
| Tone of voice | The formal register (Siz) in every sentence in the second person or the imperative. | 100% of those sentences. |
| Tags | Every HTML tag of the source, with its attributes, in the translation. | 100% of the tags. |
| Unicode character integrity | Every character intact, with no corrupted characters and no letter from another script that only looks like the right one. | 100% of the characters. |
| Locale conventions | The punctuation marks of the target language in place of the English ones. | No pass-or-fail bar. We report it as a share, because whether to localize punctuation is a matter of house style. |
| HTML entities | No HTML character code, such as the code for a quotation mark, that the source does not carry. | No added HTML entity. |
| Valid JSON | The translated JSON document still parses. | The document parses. |
| Spacing at tag boundaries | The spaces next to HTML tags stay as in the source. | 100% of the tag boundaries. |
| Advanced requirements adds a requirement for some companies and use cases | ||
| Gender neutrality | A gender-neutral translation wherever the English source leaves the gender of a person open, as in “your manager” or “the expert”. | 100% of the sentences that refer to a person. |
Pick a quality level, or your own requirements, bars and constraints, to see the shortlisted solutions that meet them in Azerbaijani, with a column for each requirement you select. Under Optimize for, choose which of them comes first: the one with the fewest total errors, the lowest cost or the shortest time. Click a column heading to sort by it.
The shortlisted solutions on 11 models, with their Intento LQA results. Total errors per 1,000 words add the weighted general translation quality errors (minor 1, major 5, critical 10) to the missed glossary terms, sentences in the wrong register, tag errors and sentences that fail gender neutrality, each counted once. Formatting counts as met or missed and does not add to the total errors. The models that did not make the shortlist have automatic scores only: the shortlist. Inconclusive: the three LQA judges agree on fewer than 90% of the sentences of the solution, so it has no general translation quality result (09-02). Gender neutrality: the Advanced requirements count the two best solutions of the language, with the gender-neutrality agent where they need it. The results of the other solutions are shown for information.
Two solutions meet every Standard requirement in Azerbaijani, and neither leads on errors, cost and time at once. The one-step solution on OpenAI GPT-5.5 has 16.5 total errors per 1,000 words and takes 74 seconds per 1,000 words. The two-step solution (translation + post-edit) on OpenAI GPT-5.6 Sol costs less, with 19.4 total errors per 1,000 words and 86 seconds per 1,000 words. So Optimize for selects OpenAI GPT-5.5 for quality and latency, and OpenAI GPT-5.6 Sol for cost.
None of the solutions that meet every Standard requirement fit a cost cap of at most $7.77 per million characters or a time limit below 74 seconds per 1,000 words, and only the one on OpenAI GPT-5.5 fits a limit of one step. The one-step solutions under that cap or limit miss general translation quality or have no conclusive Intento LQA result. The one on Google Gemini 3.5 Flash costs $4.09 per million characters and takes 18 seconds per 1,000 words, with 36.9 weighted errors per 1,000 words, too many for general translation quality.
The Advanced requirements add gender neutrality, and both solutions that meet every Standard requirement meet it without the gender-neutrality agent, with 100% of the sentences about people gender-neutral. So in Azerbaijani the Advanced requirements add nothing to their cost, time or errors.
OpenAI GPT-5.5
one-step LLM adjusted for Standard requirements
OpenAI GPT-5.6 Sol
one-step LLM adjusted for Standard requirements
OpenAI GPT-5.5
one-step LLM adjusted for Standard requirements
OpenAI GPT-5.5
one-step LLM adjusted for Standard requirements
OpenAI GPT-5.6 Sol
two-step
OpenAI GPT-5.5
one-step LLM adjusted for Standard requirements
OpenAI GPT-5.5
one-step LLM adjusted for Standard requirements
OpenAI GPT-5.6 Sol
two-step
OpenAI GPT-5.5
one-step LLM adjusted for Standard requirements
Optimize for selects one of the solutions that meet the level: the one with the fewest total errors, the lowest cost or the shortest time. Where no solution meets the level, the card shows the available solution with the fewest total errors and what it misses. Total errors count every requirement with countable errors.
| Quality level | Tier | Solutions | Simplest solution | Model | Total errors per 1,000 words | Cost per million characters | Seconds per 1,000 words |
|---|---|---|---|---|---|---|---|
| Basic | Tier 1: simple solution + backup | ★★★ | one-step | OpenAI GPT-5.6 Sol | 19.8 | $7.78 | 86 |
| Standard | Tier 1: simple solution + backup | ★ ☆ | one-step | OpenAI GPT-5.5 | 16.5 | $11.29 | 74 |
| Advanced | Tier 1: simple solution + backup | ★ ☆ | one-step | OpenAI GPT-5.5 | 16.5 | $11.29 | 74 |
One solution per level: the simplest that meets the level, the cheapest of them where several are equally simple. ★ a one-step solution, ☆ a solution with more steps or the gender-neutrality agent.
Three solutions meet the Basic requirements, all one-step, and Optimize for selects the one on OpenAI GPT-5.6 Sol for cost and the one on OpenAI GPT-5.5 for quality and latency. The Standard requirements remove the one-step solutions on OpenAI GPT-5.6 Sol and Anthropic Claude Fable 5, both on tag handling. The two-step solution (translation + post-edit) on OpenAI GPT-5.6 Sol meets them, and Optimize for selects it for cost under the Standard and Advanced requirements. The Advanced requirements remove none.
| Model | US dollars per million characters | Weighted errors per 1,000 words | Group |
|---|---|---|---|
| Google Gemma 4 | 0.16 | 56 | Misses a Standard requirement |
| z.ai GLM-5 | 2.48 | 42.5 | Misses a Standard requirement |
| Google Gemini 3.5 Flash | 4.09 | 36.9 | Misses a Standard requirement |
| Google Gemini 3.1 Pro Preview | 5.42 | 26 | Misses a Standard requirement |
| OpenAI GPT-5.6 Sol | 7.78 | 19.1 | Misses a Standard requirement |
| OpenAI GPT-5.5 | 11.29 | 16.5 | Meets every Standard requirement |
| Google Translation LLM | 21.35 | 83.1 | Misses a Standard requirement |
| DeepL Next Gen | 27.5 | 71.2 | Misses a Standard requirement |
| Anthropic Claude Fable 5 | 33.66 | 19.4 | Misses a Standard requirement |
| Requirement | Median errors per 1,000 words |
|---|---|
| General translation quality | 36.9 |
| Tag handling | 1.3 |
| Tone of voice | 0 |
| Gender neutrality | 0 |
The requirements with countable errors. Every solution with its result on each requirement: the explorer.
Of the one-step solutions that miss general translation quality, the one on Google Gemini 3.1 Pro Preview comes closest. It meets the bars for critical and major errors, with three major errors, but has 26.0 weighted errors per 1,000 words. The two-step solution (translation + post-edit) on the same model has 22.4 weighted errors per 1,000 words, still above the bar. On OpenAI GPT-5.6 Sol the post-edit step closes the gap on tags, with 100% of the tags right. Only the solutions on the LLM-based MT engines miss formatting.
| Model | Weighted errors per 1,000 words of the shortlisted one-step solutions in Azerbaijani | Class |
|---|---|---|
| Google Translation LLM | 83.1 | LLM-based MT engines |
| DeepL Next Gen | 71.2 | LLM-based MT engines |
| Google Gemma 4 | 56 | Open-weight LLMs |
| z.ai GLM-5 | 42.5 | LLMs |
| Google Gemini 3.5 Flash | 36.9 | LLMs |
| Google Gemini 3.1 Pro Preview | 26 | LLMs |
| Anthropic Claude Fable 5 | 19.4 | LLMs |
| OpenAI GPT-5.6 Sol | 19.1 | LLMs |
| OpenAI GPT-5.5 | 16.5 | LLMs |
The Intento LQA results group by model type in Azerbaijani. The six LLM solutions with a conclusive result have 16.5 to 42.5 weighted errors per 1,000 words, the solution on the open-weight Google Gemma 4 has 56.0, and those on the LLM-based MT engines DeepL Next Gen and Google Translation LLM have 71.2 and 83.1. The translations from Google Translation LLM also have 10 critical errors. All seven of these LLMs support Azerbaijani by our measured criteria, and the solutions on three of them meet general translation quality.
| Model | COMET-Kiwi | Rank on COMET-Kiwi | MetricX | Rank on MetricX | Shortlisted | Shortlist place |
|---|---|---|---|---|---|---|
| xAI Grok 4.20 (non-reasoning) | 0.799 | 10 | 2.01 | 12 | No | Not shortlisted |
| Anthropic Claude Opus 4.8 | 0.799 | 11 | 2.02 | 13 | No | Not shortlisted |
| Anthropic Claude Sonnet 4.6 | 0.799 | 12 | 1.97 | 10 | No | Not shortlisted |
| DeepSeek V3.2 | 0.798 | 13 | 2.20 | 18 | No | Not shortlisted |
| MoonshotAI Kimi K2.5 | 0.796 | 15 | 2.10 | 15 | No | Not shortlisted |
| Microsoft Translator | 0.793 | 17 | 2.07 | 14 | No | Not shortlisted |
| Anthropic Claude Sonnet 5 | 0.792 | 18 | 2.19 | 17 | No | Not shortlisted |
| Mistral Large 3 | 0.785 | 19 | 2.16 | 16 | No | Not shortlisted |
| MiniMax M2.5 | 0.783 | 20 | 2.52 | 22 | No | Not shortlisted |
| Swiss AI Apertus-8B-2509 | 0.780 | 21 | 2.57 | 23 | No | Not shortlisted |
| Anthropic Claude Haiku 4.5 | 0.776 | 22 | 2.47 | 21 | No | Not shortlisted |
| OpenAI gpt-oss-120b | 0.772 | 23 | 2.36 | 20 | No | Not shortlisted |
| Cohere Command A Translate | 0.771 | 24 | 2.63 | 24 | No | Not shortlisted |
| Mistral Medium 3.1 | 0.764 | 25 | 2.73 | 26 | No | Not shortlisted |
| Alibaba Qwen3-235B-A22B | 0.762 | 26 | 2.66 | 25 | No | Not shortlisted |
| Google Cloud Advanced Translation API | 0.757 | 27 | 3.04 | 27 | No | Not shortlisted |
| Amazon Translate | 0.743 | 28 | 3.77 | 30 | No | Not shortlisted |
| ModernMT Enterprise Edition | 0.739 | 29 | 3.05 | 28 | No | Not shortlisted |
| Yandex Cloud Translate v2 | 0.737 | 30 | 3.20 | 29 | No | Not shortlisted |
| AI21 Jamba Large 1.7 | 0.622 | 31 | 6.24 | 31 | No | Not shortlisted |
| Alibaba Machine Translation API | 0.424 | 32 | 12.01 | 32 | No | Not shortlisted |
| OpenAI GPT-5.6 Sol | 0.813 | 1 | 1.46 | 1 | Yes | Top 8 by both metrics |
| Google Gemini 3.1 Pro Preview | 0.809 | 2 | 1.74 | 4 | Yes | Top 8 by both metrics |
| Google Gemini 3.5 Flash | 0.807 | 3 | 1.59 | 2 | Yes | Top 8 by both metrics |
| OpenAI GPT-5.5 | 0.807 | 4 | 1.68 | 3 | Yes | Top 8 by both metrics |
| Anthropic Claude Fable 5 | 0.806 | 5 | 1.74 | 5 | Yes | Top 8 by both metrics |
| OpenAI GPT-4.1 | 0.804 | 6 | 1.87 | 7 | Yes | Top 8 by both metrics |
| z.ai GLM-5 | 0.802 | 7 | 1.81 | 6 | Yes | Top 8 by both metrics |
| Translated Lara 2 | 0.801 | 8 | 2.00 | 11 | Yes | Top 8 by COMET-Kiwi |
| Google Gemma 4 | 0.801 | 9 | 1.90 | 8 | Yes | Top 8 by MetricX |
| DeepL Next Gen | 0.798 | 14 | 1.96 | 9 | Yes | Real-time place |
| Google Translation LLM | 0.796 | 16 | 2.22 | 19 | Yes | Real-time place |
| Model | Solution complexity | COMET-Kiwi rank | COMET-Kiwi | MetricX rank | MetricX | Shortlist |
|---|---|---|---|---|---|---|
| OpenAI GPT-5.6 Sol | Adjusted LLM | 1 | 0.813 | 1 | 1.46 | Top 8 by both metrics |
| Google Gemini 3.1 Pro Preview | Adjusted LLM | 2 | 0.809 | 4 | 1.74 | Top 8 by both metrics |
| Google Gemini 3.5 Flash | Adjusted LLM | 3 | 0.807 | 2 | 1.59 | Top 8 by both metrics |
| OpenAI GPT-5.5 | Adjusted LLM | 4 | 0.807 | 3 | 1.68 | Top 8 by both metrics |
| Anthropic Claude Fable 5 | Adjusted LLM | 5 | 0.806 | 5 | 1.74 | Top 8 by both metrics |
| OpenAI GPT-4.1 | Adjusted LLM | 6 | 0.804 | 7 | 1.87 | Top 8 by both metrics |
| z.ai GLM-5 | Adjusted LLM | 7 | 0.802 | 6 | 1.81 | Top 8 by both metrics |
| Translated Lara 2 | Adjusted NMT | 8 | 0.801 | 11 | 2.00 | Top 8 by COMET-Kiwi |
| Google Gemma 4 | Adjusted LLM | 9 | 0.801 | 8 | 1.90 | Top 8 by MetricX |
| xAI Grok 4.20 (non-reasoning) | Adjusted LLM | 10 | 0.799 | 12 | 2.01 | Not shortlisted |
| Anthropic Claude Opus 4.8 | Adjusted LLM | 11 | 0.799 | 13 | 2.02 | Not shortlisted |
| Anthropic Claude Sonnet 4.6 | Adjusted LLM | 12 | 0.799 | 10 | 1.97 | Not shortlisted |
| DeepSeek V3.2 | Adjusted LLM | 13 | 0.798 | 18 | 2.20 | Not shortlisted |
| DeepL Next Gen | Off-the-shelf NMT | 14 | 0.798 | 9 | 1.96 | Real-time place |
| MoonshotAI Kimi K2.5 | Adjusted LLM | 15 | 0.796 | 15 | 2.10 | Not shortlisted |
| Google Translation LLM | Adjusted NMT | 16 | 0.796 | 19 | 2.22 | Real-time place |
| Microsoft Translator | Off-the-shelf NMT | 17 | 0.793 | 14 | 2.07 | Not shortlisted |
| Anthropic Claude Sonnet 5 | Adjusted LLM | 18 | 0.792 | 17 | 2.19 | Not shortlisted |
| Mistral Large 3 | Adjusted LLM | 19 | 0.785 | 16 | 2.16 | Not shortlisted |
| MiniMax M2.5 | Adjusted LLM | 20 | 0.783 | 22 | 2.52 | Not shortlisted |
| Swiss AI Apertus-8B-2509 | Off-the-shelf LLM | 21 | 0.780 | 23 | 2.57 | Not shortlisted |
| Anthropic Claude Haiku 4.5 | Adjusted LLM | 22 | 0.776 | 21 | 2.47 | Not shortlisted |
| OpenAI gpt-oss-120b | Adjusted LLM | 23 | 0.772 | 20 | 2.36 | Not shortlisted |
| Cohere Command A Translate | Adjusted LLM | 24 | 0.771 | 24 | 2.63 | Not shortlisted |
| Mistral Medium 3.1 | Adjusted LLM | 25 | 0.764 | 26 | 2.73 | Not shortlisted |
| Alibaba Qwen3-235B-A22B | Adjusted LLM | 26 | 0.762 | 25 | 2.66 | Not shortlisted |
| Google Cloud Advanced Translation API | Off-the-shelf NMT | 27 | 0.757 | 27 | 3.04 | Not shortlisted |
| Amazon Translate | Off-the-shelf NMT | 28 | 0.743 | 30 | 3.77 | Not shortlisted |
| ModernMT Enterprise Edition | Off-the-shelf NMT | 29 | 0.739 | 28 | 3.05 | Not shortlisted |
| Yandex Cloud Translate v2 | Off-the-shelf NMT | 30 | 0.737 | 29 | 3.20 | Not shortlisted |
| AI21 Jamba Large 1.7 | Adjusted LLM | 31 | 0.622 | 31 | 6.24 | Not shortlisted |
| Alibaba Machine Translation API | Off-the-shelf NMT | 32 | 0.424 | 32 | 12.01 | Not shortlisted |
| OpenAI GPT-5.6 Luna | Adjusted LLM | — | 0.806 | — | 1.79 | Same model family as OpenAI GPT-5.6 Sol |
| OpenAI GPT-5.6 Terra | Adjusted LLM | — | 0.809 | — | 1.70 | Same model family as OpenAI GPT-5.6 Sol |
| xAI Grok 4.20 (reasoning) | Adjusted LLM | — | 0.793 | — | 2.05 | Same model family as xAI Grok 4.20 (non-reasoning) |
Every model we got a translation from, as the shortlist compared it: off the shelf, without context in the request, or adjusted, with it. A model that did not make the shortlist is shown adjusted, unless off the shelf it scores higher by 0.03 or more on COMET-Kiwi. Ranks among 32 models, with each model family counted once, and the top 8 of each metric in bold. Higher is better on COMET-Kiwi, lower on MetricX. The scores rank models within this language and are not a requirement result.
First on both metrics, the one-step solution on OpenAI GPT-5.6 Sol has 19.1 weighted errors per 1,000 words, against 16.5 for the one on OpenAI GPT-5.5, fourth on COMET-Kiwi and third on MetricX. The solution on Google Gemini 3.5 Flash, third on COMET-Kiwi and second on MetricX, has 36.9 weighted errors per 1,000 words. Google Gemma 4 is on the list through MetricX alone. Of the models left off the shortlist, the nearest the cut, at rank 10, are xAI Grok 4.20 (non-reasoning) on COMET-Kiwi and Anthropic Claude Sonnet 4.6 on MetricX.
| Quality level | Cheapest solution that meets every requirement of the level | Cost per million characters |
|---|---|---|
| Basic | OpenAI GPT-5.6 Sol, one-step LLM adjusted for Standard requirements | $7.78 |
| Standard | OpenAI GPT-5.6 Sol, two-step | $7.78 |
| Advanced | OpenAI GPT-5.6 Sol, two-step | $7.78 |
| Model | US dollars per million characters | Class |
|---|---|---|
| Google Gemma 4 | $0.16 | Open-weight LLMs |
| z.ai GLM-5 | $2.48 | LLMs |
| OpenAI GPT-4.1 | $3.12 | LLMs |
| Google Gemini 3.5 Flash | $4.09 | LLMs |
| Google Gemini 3.1 Pro Preview | $5.42 | LLMs |
| OpenAI GPT-5.6 Sol | $7.78 | LLMs |
| OpenAI GPT-5.5 | $11.29 | LLMs |
| Google Translation LLM | $21.35 | LLM-based MT engines |
| Translated Lara 2 | $24.99 | LLM-based MT engines |
| DeepL Next Gen | $27.5 | LLM-based MT engines |
| Anthropic Claude Fable 5 | $33.66 | LLMs |
| Model | Tokens per character | Class |
|---|---|---|
| Anthropic Claude Opus 4.8 | 0.555 | LLMs |
| Anthropic Claude Sonnet 5 | 0.554 | LLMs |
| Anthropic Claude Fable 5 | 0.552 | LLMs |
| Mistral Large 3 | 0.545 | LLMs |
| Mistral Medium 3.1 | 0.544 | LLMs |
| AI21 Jamba Large 1.7 | 0.503 | LLMs |
| Anthropic Claude Sonnet 4.6 | 0.464 | LLMs |
| Anthropic Claude Haiku 4.5 | 0.458 | LLMs |
| Alibaba Qwen3-235B-A22B | 0.446 | Open-weight LLMs |
| DeepSeek V3.2 | 0.443 | LLMs |
| z.ai GLM-5 | 0.443 | LLMs |
| Cohere Command A Translate | 0.421 | LLMs |
| xAI Grok 4.20 (non-reasoning) | 0.42 | LLMs |
| xAI Grok 4.20 (reasoning) | 0.42 | LLMs |
| MoonshotAI Kimi K2.5 | 0.4 | LLMs |
| MiniMax M2.5 | 0.389 | LLMs |
| Google Gemini 3.1 Pro Preview | 0.376 | LLMs |
| Google Gemma 4 | 0.373 | Open-weight LLMs |
| Google Gemini 3.5 Flash | 0.36 | LLMs |
| OpenAI GPT-5.6 Sol | 0.314 | LLMs |
| OpenAI GPT-5.6 Terra | 0.314 | LLMs |
| OpenAI GPT-4.1 | 0.313 | LLMs |
| OpenAI GPT-5.6 Luna | 0.312 | LLMs |
| OpenAI GPT-5.5 | 0.311 | LLMs |
| Quality level | Fastest solution that meets every requirement of the level | Seconds per 1,000 words |
|---|---|---|
| Basic | OpenAI GPT-5.5, one-step LLM adjusted for Standard requirements | 74 |
| Standard | OpenAI GPT-5.5, one-step LLM adjusted for Standard requirements | 74 |
| Advanced | OpenAI GPT-5.5, one-step LLM adjusted for Standard requirements | 74 |
| Model | Seconds per 1,000 words | Class |
|---|---|---|
| Microsoft Translator | 0.2 | MT engines |
| Google Cloud Advanced Translation API | 0.3 | MT engines |
| ModernMT Enterprise Edition | 0.3 | MT engines |
| Google Translation LLM | 0.6 | LLM-based MT engines |
| DeepL Next Gen | 0.7 | LLM-based MT engines |
| Amazon Translate | 1.4 | MT engines |
| Translated Lara 2 | 1.5 | LLM-based MT engines |
| OpenAI gpt-oss-120b | 2.5 | Open-weight LLMs |
| Yandex Cloud Translate v2 | 5.8 | MT engines |
| OpenAI GPT-5.6 Luna | 15 | LLMs |
| Google Gemini 3.5 Flash | 18 | LLMs |
| xAI Grok 4.20 (non-reasoning) | 20 | LLMs |
| OpenAI GPT-4.1 | 23 | LLMs |
| Alibaba Machine Translation API | 29 | MT engines |
| OpenAI GPT-5.6 Terra | 37 | LLMs |
| Anthropic Claude Haiku 4.5 | 50 | LLMs |
| DeepSeek V3.2 | 53 | LLMs |
| MiniMax M2.5 | 53 | LLMs |
| Swiss AI Apertus-8B-2509 | 55 | Open-weight LLMs |
| xAI Grok 4.20 (reasoning) | 55 | LLMs |
| Mistral Medium 3.1 | 57 | LLMs |
| Cohere Command A Translate | 64 | LLMs |
| Mistral Large 3 | 65 | LLMs |
| AI21 Jamba Large 1.7 | 70 | LLMs |
| OpenAI GPT-5.5 | 74 | LLMs |
| Anthropic Claude Sonnet 5 | 82 | LLMs |
| OpenAI GPT-5.6 Sol | 86 | LLMs |
| Anthropic Claude Fable 5 | 90 | LLMs |
| Anthropic Claude Sonnet 4.6 | 93 | LLMs |
| Anthropic Claude Opus 4.8 | 104 | LLMs |
| Alibaba Qwen3-235B-A22B | 172 | Open-weight LLMs |
| Google Gemma 4 | 179 | Open-weight LLMs |
| Google Gemini 3.1 Pro Preview | 206 | LLMs |
| MoonshotAI Kimi K2.5 | 242 | LLMs |
| z.ai GLM-5 | 292 | LLMs |
Real-time models are those the shortlist counts as real-time, under 2 seconds per text across the languages we timed. The chart shows the time of each model into Azerbaijani.
The eight real-time models in Azerbaijani are four MT engines, three LLM-based MT engines and the open-weight OpenAI gpt-oss-120b. Of these, only the three LLM-based MT engines made the shortlist. The one-step LLM solutions on the shortlist take from 18 seconds per 1,000 words on Google Gemini 3.5 Flash to 292 on z.ai GLM-5. The fastest solution that meets each level is the same, the one-step solution on OpenAI GPT-5.5 at 74 seconds per 1,000 words.
| The two best solutions | Solution | Sentences about people that are gender-neutral | Meets every Advanced requirement | Cost per million characters | Seconds per 1,000 words |
|---|---|---|---|---|---|
| OpenAI GPT-5.5 | one-step | 100% | ✓ | $11.29 | 74 |
| OpenAI GPT-5.6 Sol | two-step | 100% | ✓ | $7.78 | 86 |
For tone of voice in Azerbaijani we check the formal register, with Siz as the form of address, and every shortlisted one-step solution uses it in 100% of the sentences. Azerbaijani has no grammatical gender for persons, so for gender neutrality we check that a translation avoids gendered terms and adds a gender only where the English source specifies one. A default gender is never allowed. Both of the two best solutions meet this without the gender-neutrality agent.
Leave feedback