*825This is a mandamus action. After its plenary power had expired, the trial court issued an amended order omitting a Lehmann -like finality phrase that it had included in its original order. See generally Lehmann v. Har-Con Corp. , 39 S.W.3d 191, 205-06 (Tex. 2001) (discussing finality phrases). The court of appeals directed the trial court to vacate the amended order. Here, the relator seeks a writ directing the court of appeals to vacate its opinion. For seventeen years, we have relied on Lehmann to mitigate the mischief and chaos that can arise when the prospect of appeal rears its head long after the parties believed a judgment to be final. We see no need to reduce Lehmann 's ambit, and we deny Elizondo's petition for writ of mandamus. See TEX. R. APP. P. 52.8(d).
Paul Elizondo, Cynthia Elizondo, and Eagle Fabricators, Inc. (collectively, Elizondo) hired M & O Homebuilders, Inc., Orlando Cuello, Maria De Jesus Gamez, and Texas Homebuilders, LLC (collectively, the Builders) to build a home. A cost dispute arose. Elizondo sued the Builders, asserting breach of contract, fraud, negligence, and other claims. Elizondo placed a lien on the Builders' property on the theory that the Builders had improved it using funds intended for his home. For good measure, he also applied for a temporary injunction to prevent the Builders from selling the property until the underlying litigation ended. The trial court granted the temporary injunction.
The Builders argued the lien was invalid and they filed a motion to remove it. The Builders drafted and submitted an order titled "Order on Defendants' Summary Motion to Remove Invalid Lien." The trial court signed the order, which included at the bottom of its first and only page a finality phrase that stated: "This judgment is final, disposes of all claims and all parties, and is appealable. All relief not granted herein is denied." The order left lots of relief not granted-namely, all other relief Elizondo sought against the Builders. Thirty days elapsed, marking the end of the trial court's plenary power. See TEX. R. CIV. P. 329b(d). Several weeks later, Elizondo noticed the original order had disposed of his entire case. He requested an amended order, which the trial court issued, this time omitting the finality phrase.
The Builders sought mandamus relief in the court of appeals, requesting a writ directing the trial court to vacate the amended order. The Builders argued that the original order was final, and that the amended order was void since the trial court issued it after the court's plenary power had expired. See id. ("The trial court ... has plenary power to grant a new trial or to vacate, modify, correct, or reform the judgment within thirty days after the judgment is signed."). Elizondo argued that the original order was not final, and that even if it was, the finality phrase constituted a clerical error the likes of which a trial court can modify even after its plenary power expires. See id. 329b(f) (clarifying that "the [trial] court may at any time correct a clerical error in the record of a judgment and render judgment nunc pro tunc under Rule 316" (emphasis added) ).
A divided panel conditionally granted the writ in favor of the Builders. In re M & O Homebuilders, Inc. , 516 S.W.3d 101, 110 (Tex. App.-Houston [1st Dist.] 2017, orig. proceeding). The majority reasoned that Lehmann offers two independently sufficient tests under which an order rendered *826without a conventional trial on the merits may become final. See id. at 106 (citing Lehmann , 39 S.W.3d at 205-06 ). First, an order is final if it includes a finality phrase. See Lehmann , 39 S.W.3d at 206 (suggesting as a finality phrase the statement that "[t]his judgment finally disposes of all parties and all claims and is appealable"). The parties do not dispute that the finality phrase in the original order is essentially the same as the finality phrase this Court suggested in Lehmann . Second, an order is final if it actually disposes of all claims before the trial court. See id. at 205. Importantly, in this case, the majority held that it is "necessary to review the record" only under the second test. Homebuilders , 516 S.W.3d at 106. That is, according to the majority, the finality phrase rendered the order final and the record irrelevant. See Id. Since the order was final, the majority held that the amended order retracting the finality language was an attempt to correct judicial error, and therefore void. See id. at 110.
In the dissent's view, the majority's result "[c]learly ... is not intended by Lehmann ." See Homebuilders , 516 S.W.3d at 112 (Keyes, J., dissenting). The dissent argued that the majority placed too much weight on the finality phrase, and too little weight on the trial court's later "agree[ment that] it had no intention of entering a final judgment in the case." Id. at 111. The dissent also rejected the majority's conclusion that the finality phrase made the record irrelevant. See id. at 115. Finally, the dissent wrote that Lehmann and its progeny apply only in a "summary judgment or default judgment context." Id.
Elizondo seeks a writ of mandamus directing the court of appeals to vacate its opinion. He argues that the original order was not final, and that even if it was, the trial court's inclusion of the finality phrase was a clerical error. Elizondo cites Lehmann 's statement that "whether a judicial decree is a final judgment must be determined from its language and the record in the case." See Lehmann , 39 S.W.3d at 195. In Elizondo's view, the record in this case makes clear that the trial court did not intend to render a final judgment. Elizondo also argues that the original order is ambiguous because although its title refers to removing a lien, its body includes a finality phrase. In the alternative, Elizondo asks the Court to clarify that Lehmann does not apply to the trial court's original order. He argues that applying Lehmann to the original order leads to the absurd result that Elizondo loses all of his claims against the Builders. Furthermore, he argues, that result will incentivize attorneys to sneak finality language into orders that are otherwise routine.
In response, the Builders argue that the original order was final under Lehmann , which means that the amended order improperly sought to correct judicial error rather than clerical error. The Builders concede that Lehmann allows a court of appeals to look at the record, but only "[t]o determine whether an order disposes of all pending claims and parties." Id. at 205. In other words, according to the Builders, a reviewing court may not look at the record where, as here, the order includes a finality phrase. They further contend that the original order was not ambiguous on its face, and that Elizondo's argument to the contrary depends on the record. True, the order granted more relief than the Builders requested, but the Builders urge that Lehmann specifically contemplates that possibility. The windfall the Builders' received was reversible, but it was not interlocutory. See id. at 206 ("An express adjudication of all parties and claims in a case is not interlocutory merely because the record does not afford a legal basis for the adjudication. In those circumstances, the order must be appealed and reversed.").
*827Thus, the Builders argue, the court of appeals correctly held that the trial court's original order was a final judgment.
We conclude that the court of appeals correctly applied Lehmann . Elizondo had thirty days to examine the one-page order and notice that it included a finality phrase. Even if he disagreed that the order was final, he should have treated it as though it was. See id. at 196 ("A party who is uncertain whether a judgment is final must err on the side of appealing or risk losing the right to appeal."). Had he examined the order within the thirty-day window, he could have sought an amended order or pursued an appeal. Since Elizondo waited more than thirty days to contend that the order improperly disposed of his other claims, he has lost them. Though jarring for Elizondo, this outcome reflects Lehmann 's reasoning and comports with this Court's subsequent application of Lehmann 's finality tests.
Our decision in In re Daredia offers guidance. See 317 S.W.3d 247 (Tex. 2010) (orig. proceeding) (per curiam). In that case, American Express sued a corporation and an individual to recover about $750,000 due on multiple credit-card accounts. The individual answered, but the corporation did not. American Express drafted a default judgment against the corporation, which the trial court signed. The judgment included a finality phrase. The time for appeal expired, and when American Express attempted to continue its case against the individual, the individual argued the judgment was final as to all of American Express's claims-not only those against the corporation, but also those against him. The court of appeals held that the judgment was ambiguous, and thus interlocutory. This Court reversed in a per curiam opinion reasoning that "[t]he court of appeals' holding that the [order's] failure to mention [the individual] creates an ambiguity" contradicted Lehmann . Id. at 249. The order in Daredia was "unequivocal, and therefore effective." Id.
Daredia 's reasoning decides this case, too. Although the order in Daredia bore the title "Default Judgment," the argument that the difference in title matters overlooks the crucial trait the orders share: neither mentions the claims of which it disposes. See id at 248-49. If anything, the original order in this case is more susceptible to Lehmann 's analysis because it at least mentions Elizondo; the order in Daredia did not name the individual. See id. at 248. And although in Daredia the party who authored the order suffered the harm, Daredia 's reasoning depended on the finality language rather than on which party the order benefited. See id. at 249.
The court of appeals was correct that Lehmann instructs reviewing courts to look at the record "only if the order [i]s not clear and unequivocal." See Homebuilders , 516 S.W.3d at 106. Lehmann said:
• "[W]hether a judicial decree is a final judgment must be determined from its language and the record in the case." Lehmann , 39 S.W.3d at 195 (emphasis added).
• "[W]hen there has not been a conventional trial on the merits, an order or judgment is not final for purposes of appeal unless it actually disposes of every pending claim and party or unless it clearly and unequivocally states that it finally disposes of all claims and all parties." Id. at 205 (emphasis added).
• "To determine whether an order disposes of all pending claims and parties , it may of course be necessary for the appellate court to look to the *828record in the case." Id. at 205-06 (emphasis added).
Elizondo relies on the first sentence. The Builders rely on the second and third, and those sentences more accurately convey Lehmann 's holding. While the first sentence appears in Lehmann 's discussion of finality's history, id. at 195, the second and third sentences appear within the body of Lehmann 's analysis, Id. at 205-06. By implication, the second and third sentences mean that a reviewing court confronting an order that includes a finality phrase cannot look at the record. Instead, it must take the order at face value. That makes sense. If it were otherwise, finality phrases would serve no purpose. That is, if both of Lehmann 's tests allow a reviewing court to look at the record, then a reviewing court may always look at the record. That would distill Lehmann 's joint tests into a simple rule: when there has not been a conventional trial on the merits, a court must look to the record to determine whether the judgment is final. That is not Lehmann 's rule. Had it lacked the finality phrase, the original order in this case would not have disposed of all claims and parties. However, since the original order included a finality phrase, it was clear and unequivocal.
Elizondo urges the Court to conclude that the original order is ambiguous. He contrasts the finality phrase with the original order's title, which addressed the summary removal of a lien-a proceeding he argues is not intended to be final or appealable. See TEX. PROP. CODE § 53.160(e). The trouble is that the title and the finality phrase admit of only one construction: the order (correctly) removes a lien and (incorrectly) disposes of Elizondo's other claims. Elizondo's ambiguity argument also errs in attempting to create an ambiguity from language that Lehmann describes as clear and unequivocal. See Lehmann , 39 S.W.3d at 206. The order may lack a basis in law, but it is not ambiguous. Rather, the order clearly purports to dispose of all claims and all parties. Error is not the same as ambiguity.
Nor did the court of appeals reach an absurd result. Elizondo contends that Lehmann embraced the absurdity doctrine as a limit without explicitly referring to it. Even if that were true, the absurdity doctrine does not decide this case. The reason is that the doctrine is merely a presumption against absurdity, and presumptions must yield to clear and unequivocal language. But even accepting Elizondo's view of Lehmann , the outcome here was not absurd. Elizondo argues that the order could not reasonably be read as final, and that it would therefore be absurd to conclude that it was final. To the contrary, the absurd thing would be to hold that a clear and unequivocal finality phrase does not trigger Lehmann 's directive that "[a] party who is uncertain whether a judgment is final must err on the side of appealing." Id. at 196. Elizondo also urges that the court of appeals' opinion generates absurdity by encouraging unscrupulous attorneys to sneak finality phrases into commonplace orders. The Lehmann rule actually helps on this front. No rule, alas, can altogether prevent unethical attorneys from acting unethically, but Lehmann discourages trickery by making it easier to spot.
In the dissent's view, the majority erred in reading the finality phrase as an indication of intent. Homebuilders , 516 S.W.3d at 112 (Keyes, J., dissenting). Yet Lehmann 's test holds that an order is final if it "states" that it is-not if the court intends it to be. Lehmann , 39 S.W.3d at 205. The dissent's second line of attack is that "[n]either Lehmann nor Daredia expands the power of ... finality language to establish the finality of a judgment beyond the summary judgment or default judgment context."
*829Homebuilders , 516 S.W.3d at 115. However, "when there has not been a conventional trial on the merits," Lehmann 's two-pronged test applies to "an order or judgment." Lehmann , 39 S.W.3d at 205 (emphasis added). Thus, Lehmann and its progeny apply beyond summary and default judgments. Finally, the dissent points to Lehmann 's qualification that an "order that adjudicates only the plaintiff's claims against the defendant does not adjudicate a counterclaim, cross-claim, or third party claim." Homebuilders , 516 S.W.3d at 115 (quoting Lehmann , 39 S.W.3d at 205 ). However, the only claims that either order adjudicated were Elizondo's. The original order removed the lien that Elizondo had placed on the Builders' property, but it also denied all other relief.
Our conclusion that the original order was final also decides the issue whether the amended order was void as an attempt to correct judicial error. A trial court may correct clerical errors in a judgment even after its plenary power has expired. See TEX. R. CIV. P. 329b(d), (f). But it must correct judicial errors within thirty days of judgment or not at all. See Escobar v. Escobar , 711 S.W.2d 230, 231 (Tex. 1986) ("After the trial court loses its jurisdiction over a judgment, it can correct only clerical errors in the judgment by judgment nunc pro tunc." (emphasis added) ).
The trial court's inclusion of the finality phrase in the original order constituted judicial error. It is settled that "only errors made in entering a judgment are clerical; an error in rendition is judicial." Daredia , 317 S.W.3d at 249 (citing Escobar , 711 S.W.2d at 231 ). Here, as in Daredia , the trial court signed an order that one of the parties submitted. See Daredia , 317 S.W.3d at 249. As such, the finality phrase was part of the judgment that the trial court rendered. See id. ("[P]rovisions alleged to have been inserted by mistake of the attorney nevertheless become a part of the court's judgment and therefore are judicial errors when thus rendered in writing by the court." (quoting Dikeman v. Snell , 490 S.W.2d 183, 185-86 (Tex.1973) ) ). Since the amended order sought to correct judicial error after the trial court's plenary power had expired, the amended order was void.
In sum, we conclude that the original order's finality phrase was clear, unequivocal, and neither ambiguous nor absurd. The court of appeals correctly reasoned that the finality phrase rendered the record irrelevant to determining whether the order was final. Consequently, the original order was final-"erroneous, but final." Lehmann , 39 S.W.3d at 200. The trial court's amended order was an attempt to correct judicial error beyond the period of that court's plenary power. The amended order was therefore void. Elizondo should have raised the error while the trial court still had plenary power over the case. Alternatively, Elizondo should have appealed. Having done neither, he has lost his claims. We emphasize that parties may avoid this result by following Lehmann 's admonition to "err on the side of appealing or risk losing the right to appeal." Id. at 196.
When a trial court issues an order or judgment without a conventional trial on the merits, Lehmann remains the rule for identifying whether the order or judgment is final. It is a rigid rule, but that is why it is useful. Although no rule can altogether prevent parties from using finality as a sword, the requirement of clear and unequivocal language makes offensive blows easy to spot. Blunting Lehmann 's blade would neither cushion finality's cuts nor reduce their number. Accordingly, we deny Elizondo's petition for writ of mandamus. See TEX. R. APP. P. 52.8(d).