The State of Translation Automation 2026 · Languages

English to Amharic AI translation quality, cost and speed

What it takes to automate translation from English into Amharic 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.

How good is AI translation from English to Amharic?

Key findings on English to Amharic translation

  1. 1Cost driverOn Google Gemini 3.1 Pro Preview, the solution that meets every Standard requirement in Amharic takes three steps (translation + LQA + repair) and costs $612 per million characters. The one-step solution costs $5.42 per million characters and has 38.6 weighted errors per 1,000 words, against 9.9 in three steps.See the data →
  2. 2Agent effectWith the gender-neutrality agent, the three-step solution (translation + LQA + repair) on Google Gemini 3.1 Pro Preview meets every Advanced requirement in Amharic at $1,043 per million characters, against $612 without it. Its total errors fall from 11.6 to 11.5 per 1,000 words.See the data →

Find English to Amharic translation solutions for your requirements

Pick a quality level, or your own requirements, bars and constraints, to see the shortlisted solutions that meet them in Amharic, 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. 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.

How we measure it: 11-01 · 11-02

Trade-offs in English to Amharic translation

Constraints

None of the solutions with one or two steps meet the Basic requirements in Amharic, which cover general translation quality alone. For the Standard requirements, only the three-step solution (translation + LQA + repair) on Google Gemini 3.1 Pro Preview fits a cost cap from $613 to $650 per million characters or a time limit from 317 to under 485 seconds per 1,000 words. None fit a cap of at most $612 or a limit of at most 316 seconds. None of the solutions that meet every Advanced requirement cost under $1,043 per million characters.

Agent effect

For the Advanced requirements, the gender-neutrality agent works with both three-step solutions (translation + LQA + repair). On Google Gemini 3.1 Pro Preview the cost goes from $612 to $1,043 per million characters and the time from 316 to 321 seconds per 1,000 words. On OpenAI GPT-5.5 the cost goes from $651 to $1,079 per million characters and the time from 485 to 484 seconds per 1,000 words. Total errors fall from 11.6 to 11.5 and from 18.8 to 15.8 per 1,000 words.

English to Amharic translation quality levels

Optimize for
Basic requirements☆☆

Tier 2: two or more solutions, none simple

Solutions that meet them
2 (0 one-step)
Simplest solution
three-step
Fewest errors

Google Gemini 3.1 Pro Preview
three-step

Total errors per 1,000 words
11.6
Cost per million characters
$612
Seconds per 1,000 words
316
Lowest cost

Google Gemini 3.1 Pro Preview
three-step

Total errors per 1,000 words
11.6
Cost per million characters
$612
Seconds per 1,000 words
316
Shortest time

Google Gemini 3.1 Pro Preview
three-step

Total errors per 1,000 words
11.6
Cost per million characters
$612
Seconds per 1,000 words
316
Standard requirements☆☆

Tier 2: two or more solutions, none simple

Solutions that meet them
2 (0 one-step)
Simplest solution
three-step
Fewest errors

Google Gemini 3.1 Pro Preview
three-step

Total errors per 1,000 words
11.6
Cost per million characters
$612
Seconds per 1,000 words
316
Lowest cost

Google Gemini 3.1 Pro Preview
three-step

Total errors per 1,000 words
11.6
Cost per million characters
$612
Seconds per 1,000 words
316
Shortest time

Google Gemini 3.1 Pro Preview
three-step

Total errors per 1,000 words
11.6
Cost per million characters
$612
Seconds per 1,000 words
316
Advanced requirements☆☆

Tier 2: two or more solutions, none simple

Solutions that meet them
2 (0 one-step)
Simplest solution
with the gender-neutrality agent
Fewest errors

Google Gemini 3.1 Pro Preview
three-step with the gender-neutrality agent

Total errors per 1,000 words
11.5
Cost per million characters
$1,043
Seconds per 1,000 words
321
Lowest cost

Google Gemini 3.1 Pro Preview
three-step with the gender-neutrality agent

Total errors per 1,000 words
11.5
Cost per million characters
$1,043
Seconds per 1,000 words
321
Shortest time

Google Gemini 3.1 Pro Preview
three-step with the gender-neutrality agent

Total errors per 1,000 words
11.5
Cost per million characters
$1,043
Seconds per 1,000 words
321

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 levelTierSolutionsSimplest solutionModelTotal errors per 1,000 wordsCost per million charactersSeconds per 1,000 words
BasicTier 2: two or more solutions, none simple☆☆three-stepGoogle Gemini 3.1 Pro Preview11.6$612316
StandardTier 2: two or more solutions, none simple☆☆three-stepGoogle Gemini 3.1 Pro Preview11.6$612316
AdvancedTier 2: two or more solutions, none simple☆☆with the gender-neutrality agentGoogle Gemini 3.1 Pro Preview11.5$1,043321

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.

Level step-up

Neither of the two three-step solutions (translation + LQA + repair) that meet every Basic requirement in Amharic, on Google Gemini 3.1 Pro Preview and OpenAI GPT-5.5, drops out at the Standard requirements: translations from both have 100% of the sentences in the tone of voice and 100% of the tags in place, and both meet formatting.

How we measure it: 11-01 · 11-02 · 08-09

English to Amharic translation cost against errors

Chart data
Cost and weighted errors of the one-step solutions in Amharic
ModelUS dollars per million charactersWeighted errors per 1,000 wordsGroup
Google Gemma 40.16104.5Misses a Standard requirement
Google Gemini 3.5 Flash3.9754.4Misses a Standard requirement
Google Gemini 3.1 Pro Preview5.4238.6Misses a Standard requirement
Anthropic Claude Sonnet 58.3668.2Misses a Standard requirement
Microsoft Translator10147Misses a Standard requirement
Google Translation LLM18.66105.8Misses a Standard requirement
OpenAI GPT-5.6 Sol20.8823.1Misses a Standard requirement
Anthropic Claude Opus 4.821.3955.4Misses a Standard requirement
Translated Lara 224.9981.1Misses a Standard requirement
OpenAI GPT-5.530.4134.3Misses a Standard requirement
Anthropic Claude Fable 542.4927Misses a Standard requirement

How we measure it: 11-01 · 10-10

English to Amharic translation requirements

Chart data
Errors per 1,000 words of the one-step solutions in Amharic, median, by requirement
RequirementMedian errors per 1,000 words
General translation quality55.4
Gender neutrality1.7
Tag handling0.7
Tone of voice0.3

The requirements with countable errors. Every solution with its result on each requirement: the explorer.

Bottleneck

General translation quality is also what the two-step solutions (translation + post-edit) miss in Amharic: those on OpenAI GPT-5.5 and Google Gemini 3.1 Pro Preview meet tone of voice, tag handling and formatting, but have 35.6 and 37.6 weighted errors per 1,000 words, over the bar.

How we measure it: 09-09 · 09-10 · 09-11 · 09-12 · 09-18

AI models and MT engines for English to Amharic translation

Chart data
Weighted errors per 1,000 words of the shortlisted one-step solutions in Amharic
ModelWeighted errors per 1,000 words of the shortlisted one-step solutions in AmharicClass
Microsoft Translator147MT engines
Google Translation LLM105.8LLM-based MT engines
Google Gemma 4104.5Open-weight LLMs
Translated Lara 281.1LLM-based MT engines
Anthropic Claude Sonnet 568.2LLMs
Anthropic Claude Opus 4.855.4LLMs
Google Gemini 3.5 Flash54.4LLMs
Google Gemini 3.1 Pro Preview38.6LLMs
OpenAI GPT-5.534.3LLMs
Anthropic Claude Fable 527LLMs
OpenAI GPT-5.6 Sol23.1LLMs

No translation into Amharic: AI21 Jamba Large 1.7 (garbage or near-empty), ALL-Lab (far too short, wrong language, wrong script), Swiss AI Apertus-8B-2509 (garbage or near-empty), Cohere Command A Translate (garbage or near-empty).

Model types

Translations from the seven shortlisted LLMs other than the open-weight Google Gemma 4 have 23.1 to 68.2 weighted errors per 1,000 words in one step in Amharic. Translations from the three shortlisted MT engines, two of them LLM-based, have 81.1 to 147.0 weighted errors per 1,000 words. Those from Google Translation LLM have 11 critical errors, the most on the shortlist. The three shortlisted LLMs that do not meet our language-support criteria are Google Gemini 3.5 Flash, Anthropic Claude Sonnet 5 and Google Gemma 4.

How we measure it: 00-04 · 08-15 · 23-01

How we shortlisted English to Amharic translation solutions

Chart data
Ranks of the models on the two automatic metrics, Amharic
ModelCOMET-KiwiRank on COMET-KiwiMetricXRank on MetricXShortlistedShortlist place
DeepSeek V3.20.743101.499NoNot shortlisted
Anthropic Claude Sonnet 4.60.734121.5410NoNot shortlisted
xAI Grok 4.20 (reasoning)0.726131.7514NoNot shortlisted
MoonshotAI Kimi K2.50.717151.7113NoNot shortlisted
z.ai GLM-50.715161.9116NoNot shortlisted
Google Cloud Advanced Translation API0.709172.3819NoNot shortlisted
ModernMT Enterprise Edition0.695182.2417NoNot shortlisted
OpenAI GPT-4.10.681192.2818NoNot shortlisted
MiniMax M2.50.672202.8322NoNot shortlisted
Mistral Large 30.657212.6621NoNot shortlisted
Anthropic Claude Haiku 4.50.655222.6320NoNot shortlisted
Amazon Translate0.649233.0423NoNot shortlisted
Yandex Cloud Translate v20.647243.4424NoNot shortlisted
Cohere Tiny Aya0.603254.5125NoNot shortlisted
OpenAI gpt-oss-120b0.580265.0626NoNot shortlisted
Alibaba Machine Translation API0.501276.8927NoNot shortlisted
Mistral Medium 3.10.461288.2328NoNot shortlisted
Google Gemini 3.1 Pro Preview0.75811.242YesTop 8 by both metrics
Anthropic Claude Fable 50.75821.141YesTop 8 by both metrics
OpenAI GPT-5.6 Sol0.75731.273YesTop 8 by both metrics
Google Gemini 3.5 Flash0.75141.468YesTop 8 by both metrics
Google Gemma 40.75051.335YesTop 8 by both metrics
OpenAI GPT-5.50.74961.397YesTop 8 by both metrics
Google Translation LLM0.74971.8615YesTop 8 by COMET-Kiwi
Anthropic Claude Sonnet 50.74881.314YesTop 8 by both metrics
Anthropic Claude Opus 4.80.74891.366YesTop 8 by MetricX
Translated Lara 20.740111.5411YesReal-time place
Microsoft Translator0.719141.7012YesReal-time place
ModelSolution complexityCOMET-Kiwi rankCOMET-KiwiMetricX rankMetricXShortlist
Google Gemini 3.1 Pro PreviewAdjusted LLM10.75821.24Top 8 by both metrics
Anthropic Claude Fable 5Adjusted LLM20.75811.14Top 8 by both metrics
OpenAI GPT-5.6 SolAdjusted LLM30.75731.27Top 8 by both metrics
Google Gemini 3.5 FlashAdjusted LLM40.75181.46Top 8 by both metrics
Google Gemma 4Adjusted LLM50.75051.33Top 8 by both metrics
OpenAI GPT-5.5Adjusted LLM60.74971.39Top 8 by both metrics
Google Translation LLMAdjusted NMT70.749151.86Top 8 by COMET-Kiwi
Anthropic Claude Sonnet 5Adjusted LLM80.74841.31Top 8 by both metrics
Anthropic Claude Opus 4.8Adjusted LLM90.74861.36Top 8 by MetricX
DeepSeek V3.2Adjusted LLM100.74391.49Not shortlisted
Translated Lara 2Adjusted NMT110.740111.54Real-time place
Anthropic Claude Sonnet 4.6Adjusted LLM120.734101.54Not shortlisted
xAI Grok 4.20 (reasoning)Adjusted LLM130.726141.75Not shortlisted
Microsoft TranslatorOff-the-shelf NMT140.719121.70Real-time place
MoonshotAI Kimi K2.5Adjusted LLM150.717131.71Not shortlisted
z.ai GLM-5Adjusted LLM160.715161.91Not shortlisted
Google Cloud Advanced Translation APIOff-the-shelf NMT170.709192.38Not shortlisted
ModernMT Enterprise EditionOff-the-shelf NMT180.695172.24Not shortlisted
OpenAI GPT-4.1Adjusted LLM190.681182.28Not shortlisted
MiniMax M2.5Adjusted LLM200.672222.83Not shortlisted
Mistral Large 3Adjusted LLM210.657212.66Not shortlisted
Anthropic Claude Haiku 4.5Off-the-shelf LLM220.655202.63Not shortlisted
Amazon TranslateOff-the-shelf NMT230.649233.04Not shortlisted
Yandex Cloud Translate v2Off-the-shelf NMT240.647243.44Not shortlisted
Cohere Tiny AyaOff-the-shelf LLM250.603254.51Not shortlisted
OpenAI gpt-oss-120bAdjusted LLM260.580265.06Not shortlisted
Alibaba Machine Translation APIOff-the-shelf NMT270.501276.89Not shortlisted
Mistral Medium 3.1Adjusted LLM280.461288.23Not shortlisted
OpenAI GPT-5.6 LunaAdjusted LLM—0.748—1.35Same model family as OpenAI GPT-5.6 Sol
OpenAI GPT-5.6 TerraAdjusted LLM—0.754—1.29Same model family as OpenAI GPT-5.6 Sol
xAI Grok 4.20 (non-reasoning)Adjusted LLM—0.719—1.69Same model family as xAI Grok 4.20 (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 28 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.

Metrics vs LQA

The one-step solution on Google Gemini 3.1 Pro Preview, first on COMET-Kiwi and second on MetricX, has 38.6 weighted errors per 1,000 words by Intento LQA, and the one on OpenAI GPT-5.6 Sol, third on both metrics, has 23.1. Google Translation LLM is on the shortlist through COMET-Kiwi alone, seventh there and 15th on MetricX. The one-step solution on that model has 105.8 weighted errors per 1,000 words. Nearest the cut in the tail is DeepSeek V3.2, tenth on COMET-Kiwi and ninth on MetricX.

How we measure it: 00-08 · 00-09

English to Amharic translation cost

Quality levelCheapest solution that meets every requirement of the levelCost per million characters
BasicGoogle Gemini 3.1 Pro Preview, three-step$612
StandardGoogle Gemini 3.1 Pro Preview, three-step$612
AdvancedGoogle Gemini 3.1 Pro Preview, three-step with the gender-neutrality agent$1,043
Chart data
Cost of the shortlisted one-step solutions in Amharic
ModelUS dollars per million charactersClass
Google Gemma 4$0.16Open-weight LLMs
Google Gemini 3.5 Flash$3.97LLMs
Google Gemini 3.1 Pro Preview$5.42LLMs
Anthropic Claude Sonnet 5$8.36LLMs
Microsoft Translator$10MT engines
Google Translation LLM$18.66LLM-based MT engines
OpenAI GPT-5.6 Sol$20.88LLMs
Anthropic Claude Opus 4.8$21.39LLMs
Translated Lara 2$24.99LLM-based MT engines
OpenAI GPT-5.5$30.41LLMs
Anthropic Claude Fable 5$42.49LLMs
Chart data
Tokens per character of the translation into Amharic, by model
ModelTokens per characterClass
Mistral Medium 3.11.962LLMs
z.ai GLM-51.707LLMs
Mistral Large 31.608LLMs
xAI Grok 4.20 (non-reasoning)1.566LLMs
xAI Grok 4.20 (reasoning)1.516LLMs
DeepSeek V3.21.38LLMs
OpenAI GPT-5.6 Sol1.287LLMs
OpenAI GPT-4.11.262LLMs
OpenAI GPT-5.6 Luna1.259LLMs
OpenAI GPT-5.6 Terra1.246LLMs
OpenAI GPT-5.51.238LLMs
MiniMax M2.51.137LLMs
Anthropic Claude Fable 50.991LLMs
Anthropic Claude Opus 4.80.983LLMs
Anthropic Claude Sonnet 50.977LLMs
Anthropic Claude Haiku 4.50.965LLMs
Anthropic Claude Sonnet 4.60.963LLMs
MoonshotAI Kimi K2.50.96LLMs
Cohere Tiny Aya0.56LLMs
Google Gemini 3.1 Pro Preview0.504LLMs
Google Gemma 40.494Open-weight LLMs
Google Gemini 3.5 Flash0.491LLMs

How we measure it: 10-01 · 10-03 · 05-02

English to Amharic translation speed

Quality levelFastest solution that meets every requirement of the levelSeconds per 1,000 words
BasicGoogle Gemini 3.1 Pro Preview, three-step316
StandardGoogle Gemini 3.1 Pro Preview, three-step316
AdvancedGoogle Gemini 3.1 Pro Preview, three-step with the gender-neutrality agent321
Chart data
Time to translate 1,000 words into Amharic, by model
ModelSeconds per 1,000 wordsClass
Google Cloud Advanced Translation API0.2MT engines
Microsoft Translator0.3MT engines
ModernMT Enterprise Edition0.5MT engines
Google Translation LLM1.4LLM-based MT engines
Amazon Translate1.5MT engines
Yandex Cloud Translate v22.8MT engines
Translated Lara 25.2LLM-based MT engines
OpenAI gpt-oss-120b6.7Open-weight LLMs
Alibaba Machine Translation API26MT engines
Cohere Tiny Aya38LLMs
OpenAI GPT-5.6 Luna49LLMs
DeepSeek V3.251LLMs
xAI Grok 4.20 (non-reasoning)61LLMs
Google Gemini 3.5 Flash68LLMs
Anthropic Claude Haiku 4.573LLMs
OpenAI GPT-4.175LLMs
Google Gemini 3.1 Pro Preview93LLMs
OpenAI GPT-5.6 Terra104LLMs
MiniMax M2.5124LLMs
xAI Grok 4.20 (reasoning)138LLMs
Anthropic Claude Sonnet 4.6154LLMs
Anthropic Claude Sonnet 5172LLMs
Google Gemma 4180Open-weight LLMs
Anthropic Claude Fable 5200LLMs
OpenAI GPT-5.6 Sol201LLMs
OpenAI GPT-5.5216LLMs
Mistral Large 3220LLMs
Anthropic Claude Opus 4.8226LLMs
MoonshotAI Kimi K2.5262LLMs
Mistral Medium 3.1337LLMs
z.ai GLM-5508LLMs

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 Amharic.

Model types

In Amharic, the one-step solutions take 68 to 226 seconds per 1,000 words on the eight shortlisted LLMs, against 0.3 to 5.2 on the three shortlisted MT engines.

How we measure it: 12-01 · 12-02 · 12-09

Tone of voice and gender-neutral English to Amharic translation

Tone of voice
Formal register, with respectful verb forms
Gender-neutrality strategy
neutral rephrasing
Options in the language
masculine default, double (slash) forms, neutral rephrasing (plural, nominal, infinitive)
Default gender
None where the source leaves the gender open
The two best solutionsSolutionSentences about people that are gender-neutralMeets every Advanced requirementCost per million charactersSeconds per 1,000 words
Google Gemini 3.1 Pro Previewthree-step, with the gender-neutrality agent100%✓$1,043321
OpenAI GPT-5.5three-step, with the gender-neutrality agent100%✓$1,079484
Analysis

For Amharic we check the formal register, with respectful verb forms. For gender neutrality we check neutral rephrasing through plural, nominal or infinitive forms, and a default gender is not allowed. The other options in Amharic are a masculine default and double (slash) forms. Both three-step solutions (translation + LQA + repair) meet tone of voice, and with the gender-neutrality agent 100% of the sentences about people in their translations are gender-neutral.

How we measure it: 02-05 · 02-06 · 08-11 · 09-18

How confident are we in measuring English to Amharic translation quality?

Confidence

The quality-estimation pair of automatic metrics has an agreement of 0.68 in Amharic, against a threshold of 0.53, and the similarity pair 0.57, against a threshold of 0.46. For the 11 shortlisted solutions, the median share of sentences on which the three LQA judges agree is 92.1%. With agreement at these levels, the shortlist and the Intento LQA error counts on this page carry weight.

How we measure it: 00-07 · 07-05 · 09-02

Automate translation to meet your requirements

See Intento Language Hub in action.

Book a demo
Leave feedback